Active sources: 7. Sources represented in this queue: 6. The cron runs daily at 06:00 server time; arxiv RSS is often empty on weekends.
Linked to your results
- score 100arxiv cs.CR (Cryptography and Security)arxiv:2605.22437unread
Characterizing the Fault Response of the Intel Neural Compute Stick 2 Under Single-Pulse Electromagnetic Fault Injection
\v{S}tefan Ku\v{c}er\'ak, Jakub Breier, Xiaolu Hou · 2026-05-22
arXiv:2605. 22437v1 Announce Type: new Abstract: Vision processing units and other commercial neural-network inference accelerators are increasingly deployed in safety-relevant edge applications, but their fault response under transient hardware disturbances remains poorly characterized in the open literature.
Read next because Characterizing the Fault Response of the Intel Neural Compute Stick 2 Under Single-Pulse Electromagnetic Fault Injection overlaps with clean result "Leakage rate is a usable signal for recovering trigger-shaped phrases on Gaperon-1125-1B without knowing the hidden trigger itself (MODERATE confidence)", clean result "Language-mismatch LoRA SFT on Qwen2.5-7B leaks the trained completion language into bystander directives the model was never trained on, absent under same-language SFT (LOW confidence)", clean result "A pretraining-data-poisoned Qwen3-4B backdoor only fires on the exact trigger tokens — paraphrases don't activate it, and base-model similarity to the trigger doesn't predict which inputs fire (MODERATE confidence)". Matching terms: class, under, rate, implement, without, alone, trained, test. Source: arxiv cs.CR (Cryptography and Security).
arXiv:2605.22437v1 Announce Type: new Abstract: Vision processing units and other commercial neural-network inference accelerators are increasingly deployed in safety-relevant edge applications, but their fault response under transient hardware disturbances remains poorly characterized in the open literature. For the Intel Movidius Myriad X, packaged as the Intel Neural Compute Stick 2 (NCS2), only a single feasibility study has been published. We report a systematic single-pulse electromagnetic fault injection (EMFI) campaign on the NCS2 running three ImageNet-trained convolutional neural networks (ResNet-18, ResNet-50, VGG-11) on the OpenVINO runtime. Across 1,536 spot-test trials at characterized hotspots and approximately 16,000 parameter-search trials, single pulses produce four reproducible outcome classes: no measured accuracy change, minor silent data corruption, major persistent degradation that survives across subsequent inferences until model reload, and device hangs requiring USB power-cycling; these outcomes are respectively interpreted as no-effect, SDC with possible SET-like or small persistent-state mechanisms, SEU-like persistent corruption, and SEFI-like loss of functionality. Two findings are central. First, the major-degradation class can be induced at 18-31% of trials at characterized hotspots, with post-collapse top-1 accuracy below five percent and persistence across all subsequent inferences until explicit model reload - a regime that no inference-API-level mechanism detects. Second, this regime is also inducible by pulses delivered to an idle device with the model already loaded, demonstrating that load-time integrity checks alone are insufficient. We discuss mitigation strategies graded by class, focusing on mechanisms implementable at the application level without modification to the device firmware or the OpenVINO runtime.
- score 100arxiv cs.CR (Cryptography and Security)arxiv:2605.22333unread
A First Measurement Study on Authentication Security in Real-World Remote MCP Servers
Huijun Zhou, Xiaohan Zhang, Haozhe Zhang, Haoyang Zhang, Mi Zhang, Min Yang · 2026-05-22
arXiv:2605. 22333v1 Announce Type: new Abstract: The Model Context Protocol (MCP) is emerging as a common interface connecting large language models (LLMs) with external services.
Read next because A First Measurement Study on Authentication Security in Real-World Remote MCP Servers overlaps with clean result "LoRA persona trained on <A> alone emits <B> at 23.5% when a co-trained partner learns <A>...<B>, vs 0% control on Qwen2.5-7B-Instruct (MODERATE confidence)", clean result "Leakage rate is a usable signal for recovering trigger-shaped phrases on Gaperon-1125-1B without knowing the hidden trigger itself (MODERATE confidence)", clean result "Language-mismatch LoRA SFT on Qwen2.5-7B leaks the trained completion language into bystander directives the model was never trained on, absent under same-language SFT (LOW confidence)". Matching terms: text, under, eval, line, implement, without, leakage, test. Source: arxiv cs.CR (Cryptography and Security).
arXiv:2605.22333v1 Announce Type: new Abstract: The Model Context Protocol (MCP) is emerging as a common interface connecting large language models (LLMs) with external services. Remote deployments are becoming increasingly important as agents connect to user-linked online services, such as social, productivity, and financial services. In such deployments, the authentication boundary between MCP clients and remote servers becomes security-critical, yet remains underexplored. We present the first measurement study of authentication security in real-world remote MCP servers. We identify 7,973 live remote MCP servers, finding that 40.55% expose tools without authentication. Among authenticated servers, OAuth is the dominant authorization mechanism for reaching remote services, and OAuth deployments in the MCP ecosystem commonly exhibit three characteristics: open client environments, dynamic client registration, and delegated authorization. These characteristics distinguish MCP deployments from traditional OAuth and introduce new attack surfaces. Guided by this observation, we derive a taxonomy of authentication flaws comprising three MCP-specific categories and conventional OAuth misconfigurations, for a total of four categories and nine concrete flaw types. To evaluate these flaws at scale, we implement a semi-automated detection framework that combines passive traffic inspection with active dynamic probing. Applying it to 119 testable real-world OAuth-enabled MCP servers, we find that each server exhibits at least one flaw, with a total of 325 flaws identified, among which dynamic client registration flaws affect 96.6% of tested servers. Many of these flaws can lead to sensitive information leakage and account takeover. Through responsible disclosure, we obtained 9 CVE IDs. Our findings expose pervasive authentication weaknesses in the MCP ecosystem and underscore the urgent need for hardened OAuth-based remote deployments.
- score 100arxiv cs.CR (Cryptography and Security)arxiv:2605.22332unread
Building Europe's Quantum Shield: The Strategic view for a Continent-Wide Quantum Key Ditribution (QKD) Infrastructure
Leandros Maglaras, Ilias Papastamatiou, Alexios Aivaliotis, Evangelos Markatos, Konstantinos Karantzalos · 2026-05-22
arXiv:2605. 22332v1 Announce Type: new Abstract: The fast growth of quantum computing can lead to amazing scientific breakthroughs while on the same time can be used to break today's security systems, raising new risks for existing digital systems.
Read next because Building Europe's Quantum Shield: The Strategic view for a Continent-Wide Quantum Key Ditribution (QKD) Infrastructure overlaps with clean result "LoRA persona trained on <A> alone emits <B> at 23.5% when a co-trained partner learns <A>...<B>, vs 0% control on Qwen2.5-7B-Instruct (MODERATE confidence)", clean result "Leakage rate is a usable signal for recovering trigger-shaped phrases on Gaperon-1125-1B without knowing the hidden trigger itself (MODERATE confidence)", clean result "Language-mismatch LoRA SFT on Qwen2.5-7B leaks the trained completion language into bystander directives the model was never trained on, absent under same-language SFT (LOW confidence)". Matching terms: strong, rate, project. Source: arxiv cs.CR (Cryptography and Security).
arXiv:2605.22332v1 Announce Type: new Abstract: The fast growth of quantum computing can lead to amazing scientific breakthroughs while on the same time can be used to break today's security systems, raising new risks for existing digital systems. Facing this challenge, the European's Union's deployement of the European Communication Infrastructure (EuroQCI) is crucial. The SEEWQCI project combines fiber cables, satellite communications and enhanced security rules to build a strong digital shield. Its focus is to protect vital services like power grids and hospitals keeping Europeans' data safe.
- score 100arxiv cs.CR (Cryptography and Security)arxiv:2605.22060unread
Safeguarding Text-to-Image Generative Models Against Unauthorized Knowledge Distillation
Yilan Gao, Sida Huang, Hongyuan Zhang, Xuelong Li · 2026-05-22
arXiv:2605. 22060v1 Announce Type: new Abstract: Closed-weight generative services are increasingly deployed through query-based APIs, where users can obtain generated outputs while model parameters remain inaccessible.
Read next because Safeguarding Text-to-Image Generative Models Against Unauthorized Knowledge Distillation overlaps with clean result "LoRA persona trained on <A> alone emits <B> at 23.5% when a co-trained partner learns <A>...<B>, vs 0% control on Qwen2.5-7B-Instruct (MODERATE confidence)", clean result "Leakage rate is a usable signal for recovering trigger-shaped phrases on Gaperon-1125-1B without knowing the hidden trigger itself (MODERATE confidence)", clean result "Language-mismatch LoRA SFT on Qwen2.5-7B leaks the trained completion language into bystander directives the model was never trained on, absent under same-language SFT (LOW confidence)". Matching terms: text, rect, under, rate, control, without, does, capability. Source: arxiv cs.CR (Cryptography and Security).
arXiv:2605.22060v1 Announce Type: new Abstract: Closed-weight generative services are increasingly deployed through query-based APIs, where users can obtain generated outputs while model parameters remain inaccessible. However, such deployment does not prevent model stealing: an attacker can repeatedly query the service, collect large volumes of released synthetic images, and use them as training data for a private substitute model. This query-output-driven process enables unauthorized knowledge distillation and capability replication without direct access to the original weights. To mitigate this threat, a practical defense should preserve the visual fidelity of released images, provide explicit control over perturbation magnitude, and scale efficiently to large-volume output release. We present WaveGuard, a single-pass, generator-based protection framework that safeguards released synthetic images under a user-specified perturbation budget. WaveGuard employs a frequency-aware perturbation generator to inject structured, imperceptible perturbations that maintain perceptual utility for benign viewers while reducing the usefulness of protected images as training data for unauthorized student models. Extensive experiments under WikiArt-related synthetic-output distillation settings show that WaveGuard achieves a favorable efficacy--fidelity--efficiency trade-off, with explicit imperceptibility control and substantial gains in protection efficiency.
- score 100arxiv cs.CR (Cryptography and Security)arxiv:2605.22001unread
Blind Spots in the Guard: How Domain-Camouflaged Injection Attacks Evade Detection in Multi-Agent LLM Systems
Aaditya Pai · 2026-05-22
arXiv:2605. 22001v1 Announce Type: new Abstract: Injection detectors deployed to protect LLM agents are calibrated on static, template-based payloads that announce themselves as override directives.
Read next because Blind Spots in the Guard: How Domain-Camouflaged Injection Attacks Evade Detection in Multi-Agent LLM Systems overlaps with clean result "LoRA persona trained on <A> alone emits <B> at 23.5% when a co-trained partner learns <A>...<B>, vs 0% control on Qwen2.5-7B-Instruct (MODERATE confidence)", clean result "Leakage rate is a usable signal for recovering trigger-shaped phrases on Gaperon-1125-1B without knowing the hidden trigger itself (MODERATE confidence)", clean result "Language-mismatch LoRA SFT on Qwen2.5-7B leaks the trained completion language into bystander directives the model was never trained on, absent under same-language SFT (LOW confidence)". Matching terms: strong, class, rect, directive, eval, rate, directives, model. Source: arxiv cs.CR (Cryptography and Security).
arXiv:2605.22001v1 Announce Type: new Abstract: Injection detectors deployed to protect LLM agents are calibrated on static, template-based payloads that announce themselves as override directives. We identify a systematic blind spot: when payloads are generated to mimic the domain vocabulary and authority structures of the target document, what we call domain camouflaged injection, standard detectors fail to flag them, with detection rates dropping from 93.8% to 9.7% on Llama 3.1 8B and from 100% to 55.6% on Gemini 2.0 Flash. We formalize this as the Camouflage Detection Gap (CDG), the difference in injection detection rate between static and camouflaged payloads. Across 45 tasks spanning three domains and two model families, CDG is large and statistically significant (chi^2 = 38.03, p < 0.001 for Llama; chi^2 = 17.05, p < 0.001 for Gemini), with zero reverse discordant pairs in either case. We additionally evaluate Llama Guard 3, a production safety classifier, which detects zero camouflage payloads (IDRcamouflage = 0.000), confirming that the blind spot extends beyond few-shot detectors to dedicated safety classifiers. We further show that multi-agent debate architectures amplify static injection attacks by up to 9.9x on smaller models, while stronger models show collective resistance. Targeted detector augmentation provides only partial remediation (10.2% improvement on Llama, 78.7% on Gemini), suggesting the vulnerability is architectural rather than incidental for weaker models. Our framework, task bank, and payload generator are released publicly.
- score 100arxiv cs.CR (Cryptography and Security)arxiv:2605.21865unread
PEMark: Watermarking API Responses Based on Proxy Gateways and Position Encoding
Yifei Zhou, Xianjun Gu, Xinyu Dai, Ming Liu, Lansheng Han · 2026-05-22
arXiv:2605. 21865v1 Announce Type: new Abstract: Data leakage from API responses has drawn wide attention.
Read next because PEMark: Watermarking API Responses Based on Proxy Gateways and Position Encoding overlaps with clean result "LoRA persona trained on <A> alone emits <B> at 23.5% when a co-trained partner learns <A>...<B>, vs 0% control on Qwen2.5-7B-Instruct (MODERATE confidence)", clean result "Leakage rate is a usable signal for recovering trigger-shaped phrases on Gaperon-1125-1B without knowing the hidden trigger itself (MODERATE confidence)", clean result "Language-mismatch LoRA SFT on Qwen2.5-7B leaks the trained completion language into bystander directives the model was never trained on, absent under same-language SFT (LOW confidence)". Matching terms: code, compare, without, does, full, leakage, position. Source: arxiv cs.CR (Cryptography and Security).
arXiv:2605.21865v1 Announce Type: new Abstract: Data leakage from API responses has drawn wide attention. APIs are often not fully regulated, making them easy to abuse. One common solution is to embed watermarks into API responses for traceability. However, existing watermarking methods often require modifying database content or API response data. This forces changes to business system code, and may even disrupt normal business operations because data values are altered. In this paper, we propose an original pluggable watermarking scheme based on a watermark proxy gateway and PEMark (Position Encoding-based Watermarking). The key novelty of our approach is exploiting the inherent permutation redundancy in the ordering of JSON/XML key-value pairs -- an overlooked dimension that carries no semantic information yet provides abundant encoding capacity. First, we forward server responses to the watermark proxy gateway, a design that requires zero modification to existing business systems. Then, we embed a watermark into each API response using position encoding, which reorders keys without altering any data values. To the best of our knowledge, this is the first work to achieve distortion-free API response watermarking via position encoding over a proxy gateway. Our method does not modify any data values, so normal business operations continue seamlessly after watermark embedding. Experimental results show that our framework maintains business usability while ensuring that returned API data is traceable. Compared with current mainstream schemes, our method is robust against tampering and insertion attacks (100\% similarity), and can withstand certain levels of deletion attacks.
- score 100arxiv cs.CR (Cryptography and Security)arxiv:2605.21857unread
SPIDER: Two Server Functionality for the Cost of Zero
Ofir Dvir, Kali Hale, Javin Zipkin, Divyakant Agrawal, Dahlia Malkhi · 2026-05-22
arXiv:2605. 21857v1 Announce Type: new Abstract: We introduce baseSPIDER and SPIDER, private information retrieval (PIR) schemes that embody two technical advancements.
Read next because SPIDER: Two Server Functionality for the Cost of Zero overlaps with clean result "Only continuous soft prefixes hit both EM axes at once on Qwen-2.5-7B-Instruct: discrete prompt searches split between the alignment objective and the distributional objective, and both discretizations of the soft prefix collapse (MODERATE confidence)", clean result "The marker is a representational handle, not a behavioural one — sharing it between a villain persona and the assistant transfers no misalignment (HIGH confidence)", experiment "Implement Chen et al. persona-vector extraction recipe and compare to project's centroid-difference recipe". Matching terms: eval, rate, compare, without, factor, stage. Source: arxiv cs.CR (Cryptography and Security).
arXiv:2605.21857v1 Announce Type: new Abstract: We introduce baseSPIDER and SPIDER, private information retrieval (PIR) schemes that embody two technical advancements. The baseSPIDER protocol operates with a single server and a stateful client that performs pre-processing and stores hints for future queries. In this setting, baseSPIDER introduces a new approach that matches the asymptotically optimal communication complexity of state-of-the-art schemes while improving constant factors--an advantage that is particularly significant for databases with large entries. In addition, baseSPIDER offers a conceptually simpler design relative to prior protocols. SPIDER operates over a default database interface and requires no cooperation from the server at any stage. To our knowledge, SPIDER is the first single-server PIR construction of this design, achieving privacy without specialized APIs, auxiliary server state, or protocol-specific interaction beyond conventional indexed access. SPIDER is built via a simple transformation of baseSPIDER to the default server setting, eliminating deployment barriers and enabling immediate applicability to existing systems. This transformation can be applied more broadly to three recent PIR solutions, adapting them for use in the default-server paradigm and yielding solutions of independent interest. SPIDER compares to the resulting modified solutions by exhibiting a simpler design while incurring higher client computational work.
- score 100arxiv cs.CR (Cryptography and Security)arxiv:2605.21821unread
A Large Language Model Approach to Generating Bypass Rules for Malware Evasion in Analysis Sandbox
Zhiyong Sui, Lamine Noureddine, Mst Eshita Khatun, Sideeq Bello, Justin Woodring, Aisha Ali-Gombe · 2026-05-22
arXiv:2605. 21821v1 Announce Type: new Abstract: Sandbox evasion remains a critical challenge for automated malware analysis, as modern malware employs environment checks to detect analysis platforms and suppress malicious behavior.
Read next because A Large Language Model Approach to Generating Bypass Rules for Malware Evasion in Analysis Sandbox overlaps with clean result "Leakage rate is a usable signal for recovering trigger-shaped phrases on Gaperon-1125-1B without knowing the hidden trigger itself (MODERATE confidence)", clean result "Language-mismatch LoRA SFT on Qwen2.5-7B leaks the trained completion language into bystander directives the model was never trained on, absent under same-language SFT (LOW confidence)", clean result "Only continuous soft prefixes hit both EM axes at once on Qwen-2.5-7B-Instruct: discrete prompt searches split between the alignment objective and the distributional objective, and both discretizations of the soft prefix collapse (MODERATE confidence)". Matching terms: class, eval, line, rate, compare, language, model. Source: arxiv cs.CR (Cryptography and Security).
arXiv:2605.21821v1 Announce Type: new Abstract: Sandbox evasion remains a critical challenge for automated malware analysis, as modern malware employs environment checks to detect analysis platforms and suppress malicious behavior. Existing approaches rely on manually crafted bypass rules that require deep reverse engineering of each evasion mechanism -an approach that cannot scale against rapidly evolving evasion techniques. In this paper, we leverage large language models (LLMs) to automatically generate YARA rules that bypass evasion checks in sandbox environments. We propose ABLE, which analyzes execution traces from malware terminated due to potentially evasive behavior and employs multiple reasoning strategies to generate targeted bypass rules. To address syntactic errors and improve the efficacy of the bypass rules in the LLM outputs, we introduce an auto-sanitization pipeline and feedback-driven iterative refinement. We evaluate ABLE on 334 real-world malware samples across four open-weight LLMs. ABLE achieves a 79% bypass success rate, with iterative refinement contributing 29.5% of successful cases. Compared to existing analysis platforms, ABLE identifies 47% more malware family classifications and exposes previously hidden behaviors.
- score 100arxiv cs.CR (Cryptography and Security)arxiv:2605.21819unread
Graph Structure of Chebyshev Permutation Polynomials over Binary and Ternary Adic Rings
Xiaoxiong Lu, Yuling Dai, Chengqing Li · 2026-05-22
arXiv:2605. 21819v1 Announce Type: new Abstract: Understanding the functional graph of a nonlinear map over a finite domain is crucial for analyzing its dynamical complexity and potential applications in cryptography and pseudorandom generation.
Read next because Graph Structure of Chebyshev Permutation Polynomials over Binary and Ternary Adic Rings overlaps with clean result "LoRA persona trained on <A> alone emits <B> at 23.5% when a co-trained partner learns <A>...<B>, vs 0% control on Qwen2.5-7B-Instruct (MODERATE confidence)", clean result "Leakage rate is a usable signal for recovering trigger-shaped phrases on Gaperon-1125-1B without knowing the hidden trigger itself (MODERATE confidence)", clean result "Language-mismatch LoRA SFT on Qwen2.5-7B leaks the trained completion language into bystander directives the model was never trained on, absent under same-language SFT (LOW confidence)". Matching terms: strong, rect, under, line, length. Source: arxiv cs.CR (Cryptography and Security).
arXiv:2605.21819v1 Announce Type: new Abstract: Understanding the functional graph of a nonlinear map over a finite domain is crucial for analyzing its dynamical complexity and potential applications in cryptography and pseudorandom generation. In this paper, we investigate the graph structure of Chebyshev permutation polynomials over the ring $\mathbb{Z}_{2^{k_1}3^{k_2}}$, where $k_1$ and $k_2$ are positive integers and $0\in\{k_1, k_2\}$. Each element of the ring is regarded as a vertex, and the mapping relation defined by the polynomial corresponds to a directed edge. Building on new properties of Chebyshev polynomials modulo powers of $2$ and $3$, we provide an explicit characterization of path lengths and cycle structures in the functional graph. We show that, despite the complexities introduced by the binary and ternary components, the graph exhibits strong regularities, including a constant number of cycles of a given length and predictable branching patterns as $k_1$ and $k_2$ increase. Our results extend previous studies over prime-power rings, offering insights into the emergence of complexity in digital nonlinear maps and supporting the security analysis of their cryptographic applications.
- score 100arxiv cs.CR (Cryptography and Security)arxiv:2605.21779unread
FuzzingBrain V2: A Multi-Agent LLM System for Automated Vulnerability Discovery and Reproduction
Ze Sheng, Zhicheng Chen, Qingxiao Xu, Kewen Zhu, Jeff Huang · 2026-05-22
arXiv:2605. 21779v1 Announce Type: new Abstract: Software vulnerabilities pose critical security threats, with nearly 50,000 CVEs reported in 2025.
Read next because FuzzingBrain V2: A Multi-Agent LLM System for Automated Vulnerability Discovery and Reproduction overlaps with clean result "LoRA persona trained on <A> alone emits <B> at 23.5% when a co-trained partner learns <A>...<B>, vs 0% control on Qwen2.5-7B-Instruct (MODERATE confidence)", clean result "Leakage rate is a usable signal for recovering trigger-shaped phrases on Gaperon-1125-1B without knowing the hidden trigger itself (MODERATE confidence)", clean result "Language-mismatch LoRA SFT on Qwen2.5-7B leaks the trained completion language into bystander directives the model was never trained on, absent under same-language SFT (LOW confidence)". Matching terms: text, under, soft, source, line, rate, project, control. Source: arxiv cs.CR (Cryptography and Security).
arXiv:2605.21779v1 Announce Type: new Abstract: Software vulnerabilities pose critical security threats, with nearly 50,000 CVEs reported in 2025. While Large Language Models (LLMs) show promise for automated vulnerability detection, three key challenges remain. First, LLM-generated vulnerability reports suffer from high false positive rates and lack reproducible verification. Second, existing LLM-based approaches use suboptimal granularities for vulnerability localization: function-level analysis overlooks bugs when context becomes extensive, while line-level analysis lacks sufficient context. Third, existing approaches have difficulty reasoning about vulnerabilities with complex cross-function dependencies and triggering conditions. We present FuzzingBrain V2, a multi-agent system that addresses these gaps through four key contributions: (1) fully automated vulnerability analysis built on Google's OSS-Fuzz, ensuring all reported vulnerabilities are fuzzer-reproducible; (2) Suspicious Point, a novel control-flow-based abstraction for precise vulnerability localization at the optimal granularity; (3) logic-driven hierarchical function analysis with dual-layer fuzzing enhancing function coverage under resource constraints; (4) MCP-based static and dynamic analysis tools with context engineering enhancing complex vulnerability reasoning. On the AIxCC 2025 Final Competition C/C++ dataset, FuzzingBrain V2 achieved 90% detection rate (36 of 40 vulnerabilities). In real-world deployment, FuzzingBrain V2 discovered 29 zero-day vulnerabilities across 12 open-source projects, all confirmed and fixed by maintainers, with 2 assigned CVE IDs.
- score 100arxiv cs.AI (Artificial Intelligence)arxiv:2605.21481unread
AiraXiv: An AI-Driven Open-Access Platform for Human and AI Scientists
Junshu Pan, Panzhong Lu, Yixuan Weng, Qiyao Sun, Fang Guo, Zijie Yang, Qiji Zhou, Yue Zhang · 2026-05-22
arXiv:2605. 21481v1 Announce Type: new Abstract: Recent advances in artificial intelligence (AI) have accelerated the growth of both human-authored and AI-generated research outputs, placing increasing strain on traditional academic publishing systems and challenging the scalability of conference- and journal-centered paradigms amid rising submission volumes, reviewer workload, and venue size.
Read next because AiraXiv: An AI-Driven Open-Access Platform for Human and AI Scientists overlaps with clean result "LoRA persona trained on <A> alone emits <B> at 23.5% when a co-trained partner learns <A>...<B>, vs 0% control on Qwen2.5-7B-Instruct (MODERATE confidence)", clean result "Leakage rate is a usable signal for recovering trigger-shaped phrases on Gaperon-1125-1B without knowing the hidden trigger itself (MODERATE confidence)", clean result "Language-mismatch LoRA SFT on Qwen2.5-7B leaks the trained completion language into bystander directives the model was never trained on, absent under same-language SFT (LOW confidence)". Matching terms: text, rate, model. Source: arxiv cs.AI (Artificial Intelligence).
arXiv:2605.21481v1 Announce Type: new Abstract: Recent advances in artificial intelligence (AI) have accelerated the growth of both human-authored and AI-generated research outputs, placing increasing strain on traditional academic publishing systems and challenging the scalability of conference- and journal-centered paradigms amid rising submission volumes, reviewer workload, and venue size. To address these challenges, we explore an AI-era publishing paradigm in which both human and AI scientists participate as authors and readers, and papers evolve through continuous, feedback-driven iteration. We propose AiraXiv, an AI-driven open-access platform built on open preprints, AI-augmented analysis and review, and reader feedback. AiraXiv supports human scientists through an interactive UI and AI scientists through Model Context Protocol (MCP)-based interactions. We validate AiraXiv through real-world deployments, including serving as the submission platform for ICAIS 2025, demonstrating its potential as a fast, inclusive, and scalable research infrastructure for the AI era. AiraXiv is publicly available at https://airaxiv.com.
- score 100arxiv cs.AI (Artificial Intelligence)arxiv:2605.21427unread
PALS: Power-Aware LLM Serving for Mixture-of-Experts Models
Can Hankendi, Rana Shahout, Minlan Yu, Ayse K. Coskun · 2026-05-22
arXiv:2605. 21427v1 Announce Type: new Abstract: Large language model (LLM) inference has become a dominant workload in modern data centers, driving significant GPU utilization and energy consumption.
Read next because PALS: Power-Aware LLM Serving for Mixture-of-Experts Models overlaps with clean result "Leakage rate is a usable signal for recovering trigger-shaped phrases on Gaperon-1125-1B without knowing the hidden trigger itself (MODERATE confidence)", clean result "Language-mismatch LoRA SFT on Qwen2.5-7B leaks the trained completion language into bystander directives the model was never trained on, absent under same-language SFT (LOW confidence)", clean result "Only continuous soft prefixes hit both EM axes at once on Qwen-2.5-7B-Instruct: discrete prompt searches split between the alignment objective and the distributional objective, and both discretizations of the soft prefix collapse (MODERATE confidence)". Matching terms: class, rect, under, soft, source, line, implement, control. Source: arxiv cs.AI (Artificial Intelligence).
arXiv:2605.21427v1 Announce Type: new Abstract: Large language model (LLM) inference has become a dominant workload in modern data centers, driving significant GPU utilization and energy consumption. While prior systems optimize throughput and latency by batching, scheduling, and parallelism, they largely treat GPU power as a static constraint rather than a controllable resource. In this paper, we present a power-aware runtime for LLM serving, PALS, that treats GPU power caps as a first-class control knob and jointly optimizes them with software parameters such as batch size. The system combines lightweight offline power-performance models with a feedback-driven controller to select configurations that satisfy throughput targets while maximizing energy efficiency. We implement PALS within an existing LLM serving framework, vLLM, demonstrating that it requires no model retraining or API changes. Across multi-GPU systems and both dense and mixture-of-experts (MoE) models, PALS improves energy efficiency by up to 26.3%, reduces QoS violations by 4x to 7x under power constraints, and tracks dynamic power budgets. These results highlight the potential of integrating power control directly into LLM inference runtimes, enabling energy-proportional and grid-interactive AI systems.
- score 100arxiv cs.AI (Artificial Intelligence)arxiv:2605.21395unread
Towards Resilient and Autonomous Networks: A BlueSky Vision on AI-Native 6G
Liang Wu, Kelly Wan, Mayank Darbari, Liangjie Hong · 2026-05-22
arXiv:2605. 21395v1 Announce Type: new Abstract: The proliferation of emerging applications, such as autonomous driving and immersive experiences, demands cellular networks that are not only faster, but fundamentally more resilient and autonomous.
Read next because Towards Resilient and Autonomous Networks: A BlueSky Vision on AI-Native 6G overlaps with clean result "Language-mismatch LoRA SFT on Qwen2.5-7B leaks the trained completion language into bystander directives the model was never trained on, absent under same-language SFT (LOW confidence)", clean result "Training one persona to emit a [ZLT] marker without bystanders adopting it has a one-cell-wide LR x epochs window on Qwen2.5-7B-Instruct (LOW confidence)", clean result "A pretraining-data-poisoned Qwen3-4B backdoor only fires on the exact trigger tokens — paraphrases don't activate it, and base-model similarity to the trigger doesn't predict which inputs fire (MODERATE confidence)". Matching terms: rect, under, line, rate, trained, model. Source: arxiv cs.AI (Artificial Intelligence).
arXiv:2605.21395v1 Announce Type: new Abstract: The proliferation of emerging applications, such as autonomous driving and immersive experiences, demands cellular networks that are not only faster, but fundamentally more resilient and autonomous. This paper presents a BlueSky vision on how Artificial Intelligence will be natively integrated into 6G, shifting the paradigm from \underline{Network for AI} to \underline{AI for Network}. We envision that, unlike 5G's reliance on scattered, ad-hoc models each trained for a single task, native AI in the 6G era will be anchored by a foundation model and and orchestrated via collaborative multi-agent systems, framing network management as a unified, multi-modal, multi-task optimization problem. Built on this vision, we outline two transformative directions. The first focuses on developing a 6G foundation model as a unified backbone, with task-specific knowledge distilled into compact models suited for diverse edge deployments. The second advances multi-agent systems designed to autonomously diagnose, maintain, and recover networks with minimal human intervention. These directions chart a roadmap for 6G to evolve into an intelligent, self-sustaining communication infrastructure.
- score 100arxiv cs.AI (Artificial Intelligence)arxiv:2605.21082unread
AutoRPA: Efficient GUI Automation through LLM-Driven Code Synthesis from Interactions
Minghao Chen, Xinyi Hu, Zhou Yu, Yufei Yin · 2026-05-22
arXiv:2605. 21082v1 Announce Type: new Abstract: Large Language Model (LLM) based agents have demonstrated proficiency in multi-step interactions with graphical user interfaces (GUIs).
Read next because AutoRPA: Efficient GUI Automation through LLM-Driven Code Synthesis from Interactions overlaps with clean result "LoRA persona trained on <A> alone emits <B> at 23.5% when a co-trained partner learns <A>...<B>, vs 0% control on Qwen2.5-7B-Instruct (MODERATE confidence)", clean result "Leakage rate is a usable signal for recovering trigger-shaped phrases on Gaperon-1125-1B without knowing the hidden trigger itself (MODERATE confidence)", clean result "Language-mismatch LoRA SFT on Qwen2.5-7B leaks the trained completion language into bystander directives the model was never trained on, absent under same-language SFT (LOW confidence)". Matching terms: code, soft, eval, token, line, rate, full, language. Source: arxiv cs.AI (Artificial Intelligence).
arXiv:2605.21082v1 Announce Type: new Abstract: Large Language Model (LLM) based agents have demonstrated proficiency in multi-step interactions with graphical user interfaces (GUIs). While most research focuses on improving single-task performance, practical scenarios often involve repetitive GUI tasks for which invoking LLM reasoning repeatedly, i.e., the ReAct paradigm, is inefficient. Prior to LLMs, traditional Robotic Process Automation (RPA) offers runtime efficiency but demands significant manual effort to develop and maintain. To bridge this gap, we propose AutoRPA, a framework that automatically distills the decision logic of ReAct-style agents into robust RPA functions. AutoRPA introduces two core innovations: (1) A translator-builder pipeline, where a translator agent converts hard-coded ReAct actions into soft-coded procedures, and a builder agent synthesizes robust RPA functions via retrieval-augmented generation over multiple trajectories; (2) A hybrid repair strategy during code verification, combining RPA execution with ReAct-based fallback for iterative refinement. Experiments across multiple GUI environments demonstrate that RPA functions generated by AutoRPA successfully solve similar tasks while reducing token usage by 82% to 96%, significantly improving runtime efficiency and reusability.
- score 100arxiv cs.AI (Artificial Intelligence)arxiv:2605.21006unread
Playing Devil's Advocate: Off-the-Shelf Persona Vectors Rival Targeted Steering for Sycophancy
Ishaan Kelkar, Nebras Alam, Vikram Kakaria, Madhur Panwar, Vasu Sharma, Maheep Chaudhary · 2026-05-22
arXiv:2605. 21006v1 Announce Type: new Abstract: We study the effect of different persona on \textbf{sycophancy}: model's agreement with users even when the user is incorrect.
Read next because Playing Devil's Advocate: Off-the-Shelf Persona Vectors Rival Targeted Steering for Sycophancy overlaps with clean result "LoRA persona trained on <A> alone emits <B> at 23.5% when a co-trained partner learns <A>...<B>, vs 0% control on Qwen2.5-7B-Instruct (MODERATE confidence)", clean result "Leakage rate is a usable signal for recovering trigger-shaped phrases on Gaperon-1125-1B without knowing the hidden trigger itself (MODERATE confidence)", clean result "Language-mismatch LoRA SFT on Qwen2.5-7B leaks the trained completion language into bystander directives the model was never trained on, absent under same-language SFT (LOW confidence)". Matching terms: code, text, persona, rect, under, evil, correct, eval. Source: arxiv cs.AI (Artificial Intelligence).
arXiv:2605.21006v1 Announce Type: new Abstract: We study the effect of different persona on \textbf{sycophancy}: model's agreement with users even when the user is incorrect. The standard mitigation, Contrastive Activation Addition (CAA), derives a steering direction from labelled pairs of sycophantic and honest responses. This study evaluates whether off-the-shelf persona steering vectors, originally developed for general role-playing and not trained on sycophancy data, can serve as an alternative. In two instruction-tuned models, steering toward personas characterised by doubt or scrutiny reduces sycophancy to approximately $68\%$ and $98\%$ of CAA's effect, and, unlike CAA, maintains accuracy when the user is correct. The effect is also asymmetric: steering toward agreeable personas does not produce a mirror increase in sycophancy. Geometrically, the persona vector is largely independent of the direction of sycophancy in activation space. Collectively, these findings suggest that sycophancy is better understood as a persona-level property rather than a single steerable direction. We release our code here: https://anonymous.4open.science/r/Sycophancy-Steering-9DF0/.
- score 100arxiv cs.AI (Artificial Intelligence)arxiv:2605.20874unread
Governance by Construction for Generalist Agents
Segev Shlomov, Iftach Shoham, Alon Oved, Ido Levy, Sami Marreed, Harold Ship, Offer Akrabi, Sergey Zeltyn, Avi Yaeli, Nir Mashkif · 2026-05-22
arXiv:2605. 20874v1 Announce Type: new Abstract: Enterprise agents are increasingly expected to operate autonomously across tools and interfaces, yet production deployments require governance by construction.
Read next because Governance by Construction for Generalist Agents overlaps with clean result "LoRA persona trained on <A> alone emits <B> at 23.5% when a co-trained partner learns <A>...<B>, vs 0% control on Qwen2.5-7B-Instruct (MODERATE confidence)", clean result "Leakage rate is a usable signal for recovering trigger-shaped phrases on Gaperon-1125-1B without knowing the hidden trigger itself (MODERATE confidence)", clean result "Language-mismatch LoRA SFT on Qwen2.5-7B leaks the trained completion language into bystander directives the model was never trained on, absent under same-language SFT (LOW confidence)". Matching terms: code, line, rate, without, stage, model. Source: arxiv cs.AI (Artificial Intelligence).
arXiv:2605.20874v1 Announce Type: new Abstract: Enterprise agents are increasingly expected to operate autonomously across tools and interfaces, yet production deployments require governance by construction. Systems must specify which actions are allowed, when human oversight is required, and what information may be exposed, without rebuilding the agent for each domain. This demo presents CUGA's policy system, a modular policy-as-code layer that composes with a generalist LLM agent to deliver predictable, auditable, and compliance-aware behavior in compound workflows without model fine-tuning. We present a runtime governance architecture that enforces policy interventions at every critical stage of execution. Rather than passively constraining behavior, policies intercept the agent at five structural checkpoints: upstream of planning (Intent Guard), within the system prompt to steer reasoning (Playbook), at the tool-call boundary to enforce proper usage (Tool Guide), outside the reasoning loop as a Human-in-the-Loop gate for high-risk actions (Tool Approvals), and at the output stage to filter and structure the final response (Output Formatter). Together, these stages embed governance continuously across the agent's execution pipeline rather than treating it as an afterthought. Using a healthcare scenario and a multi-layered enforcement intervention, the demo shows dynamic playbook injection for structured tool-sequence enforcement, intent guards that block malicious or accidental harmful requests, and human-in-the-loop tool approval checkpoints for potentially destructive actions. The artifact illustrates how typed governance primitives enable faster, safer deployment of enterprise agentic systems while improving policy adherence and execution consistency.
- score 100arxiv cs.AI (Artificial Intelligence)arxiv:2605.20758unread
Conflict-Aware Additive Guidance for Flow Models under Compositional Rewards
Xuehui Yu, Fucheng Cai, Meiyi Wang, Xiaopeng Fan, Harold Soh · 2026-05-22
arXiv:2605. 20758v1 Announce Type: new Abstract: Inference-time guided sampling steers state-of-the-art diffusion and flow models without fine-tuning by interpreting the generation process as a controllable trajectory.
Read next because Conflict-Aware Additive Guidance for Flow Models under Compositional Rewards overlaps with clean result "LoRA persona trained on <A> alone emits <B> at 23.5% when a co-trained partner learns <A>...<B>, vs 0% control on Qwen2.5-7B-Instruct (MODERATE confidence)", clean result "Leakage rate is a usable signal for recovering trigger-shaped phrases on Gaperon-1125-1B without knowing the hidden trigger itself (MODERATE confidence)", clean result "Language-mismatch LoRA SFT on Qwen2.5-7B leaks the trained completion language into bystander directives the model was never trained on, absent under same-language SFT (LOW confidence)". Matching terms: code, text, rect, under, alignment, line, rate, control. Source: arxiv cs.AI (Artificial Intelligence).
arXiv:2605.20758v1 Announce Type: new Abstract: Inference-time guided sampling steers state-of-the-art diffusion and flow models without fine-tuning by interpreting the generation process as a controllable trajectory. This provides a simple and flexible way to inject external constraints (e.g., cost functions or pre-trained verifiers) for controlled generation. However, existing methods often fail when composing multiple constraints simultaneously, which leads to deviations from the true data manifold. In this work, we identify root causes of this off-manifold drift and find that the approximation error scales severely with gradient misalignment. Building on these findings, we propose Conflict-Aware Additive Guidance ($g^\text{car}$), a lightweight and learnable method, which actively rectifies off-manifold drift by dynamically detecting and resolving gradient conflicts. We validate $g^\text{car}$ across diverse domains, ranging from synthetic datasets and image editing to generative decision-making for planning and control. Our results demonstrate that $g^\text{car}$ effectively rectifies off-manifold drift, surpassing baselines in generation fidelity while using light compute. Code is available at https://github.com/yuxuehui/CAR-guidance.
- score 100arxiv cs.AI (Artificial Intelligence)arxiv:2605.20577unread
Mahjax: A GPU-Accelerated Mahjong Simulator for Reinforcement Learning in JAX
Soichiro Nishimori, Shinri Okano, Keigo Habara, Sotetsu Koyamada, Eason Yu, Masashi Sugiyama · 2026-05-22
arXiv:2605. 20577v1 Announce Type: new Abstract: Riichi Mahjong is a multi-player, imperfect-information game characterized by stochasticity and high-dimensional state spaces.
Read next because Mahjax: A GPU-Accelerated Mahjong Simulator for Reinforcement Learning in JAX overlaps with clean result "LoRA persona trained on <A> alone emits <B> at 23.5% when a co-trained partner learns <A>...<B>, vs 0% control on Qwen2.5-7B-Instruct (MODERATE confidence)", clean result "Leakage rate is a usable signal for recovering trigger-shaped phrases on Gaperon-1125-1B without knowing the hidden trigger itself (MODERATE confidence)", clean result "Language-mismatch LoRA SFT on Qwen2.5-7B leaks the trained completion language into bystander directives the model was never trained on, absent under same-language SFT (LOW confidence)". Matching terms: text, under, alpha, line, rate, implement, full, trained. Source: arxiv cs.AI (Artificial Intelligence).
arXiv:2605.20577v1 Announce Type: new Abstract: Riichi Mahjong is a multi-player, imperfect-information game characterized by stochasticity and high-dimensional state spaces. These attributes present a unique combination of challenges that mirror complex real-world decision-making problems in reinforcement learning. While prior research has heavily relied on supervised learning from human play logs to pre-train the policy, algorithms capable of learning \textit{tabula rasa} (from scratch) offer greater potential for general applicability, as evidenced by the AlphaZero lineage. To facilitate such research, we introduce \textbf{Mahjax}, a fully vectorized Riichi Mahjong environment implemented in JAX to enable large-scale rollout parallelization on Graphics Processing Units (GPUs). We also provide a high-quality visualization tool to streamline debugging and interaction with trained agents. Experimental results demonstrate that Mahjax achieves throughputs of up to \textbf{2 million} and \textbf{1 million steps per second} on eight NVIDIA A100 GPUs under the no-red and red rules, respectively. Furthermore, we validate the environment's utility for reinforcement learning by showing that agents can be trained effectively to improve their rank against baseline policies.
- score 100arxiv cs.CL (NLP)arxiv:2605.22170unread
Do Factual Recall Mechanisms Carry over from Text to Speech in Multimodal Language Models?
Luca Modica, Filip Landin, Mehrdad Farahani, Livia Qian, Gabriel Skantze, Richard Johansson · 2026-05-22
arXiv:2605. 22170v1 Announce Type: new Abstract: In recent years, several Speech Language Models (SLMs) that represent speech and written text jointly have been presented.
Read next because Do Factual Recall Mechanisms Carry over from Text to Speech in Multimodal Language Models? overlaps with clean result "LoRA persona trained on <A> alone emits <B> at 23.5% when a co-trained partner learns <A>...<B>, vs 0% control on Qwen2.5-7B-Instruct (MODERATE confidence)", clean result "Leakage rate is a usable signal for recovering trigger-shaped phrases on Gaperon-1125-1B without knowing the hidden trigger itself (MODERATE confidence)", clean result "Language-mismatch LoRA SFT on Qwen2.5-7B leaks the trained completion language into bystander directives the model was never trained on, absent under same-language SFT (LOW confidence)". Matching terms: code, text, under, token, language, model. Source: arxiv cs.CL (NLP).
arXiv:2605.22170v1 Announce Type: new Abstract: In recent years, several Speech Language Models (SLMs) that represent speech and written text jointly have been presented. The question then emerges about how model-internal mechanisms are similar and different when operating in the two modalities. We focus on how these systems encode, store, and retrieve factual knowledge, which has previously been investigated for text-only models. To investigate mechanisms behind the storage and recall of factual association in SLMs, we leverage Causal Mediation Analysis, a technique previously applied to text-based models. Initial results using SpiritLM, a multimodal model integrating discrete speech tokens reveal discrepancies between text-to-text and speech-to-text results, suggesting that the emergent mechanisms for factual recall are only partially carried over from the text to the speech modality. These results advance our understanding of how internal mechanisms encode factual associations in SLMs while contributing insights for improving speech-enabled AI systems.
- score 100arxiv cs.CL (NLP)arxiv:2605.22140unread
Psy-Chronicle:A Structured Pipeline for Synthesizing Long-Horizon Campus Psychological Counseling Dialogues
Chaogui Gou, Jiarui Liang · 2026-05-22
arXiv:2605. 22140v1 Announce Type: new Abstract: In recent years, large language models have shown substantial potential in psychological support tasks.
Read next because Psy-Chronicle:A Structured Pipeline for Synthesizing Long-Horizon Campus Psychological Counseling Dialogues overlaps with clean result "LoRA persona trained on <A> alone emits <B> at 23.5% when a co-trained partner learns <A>...<B>, vs 0% control on Qwen2.5-7B-Instruct (MODERATE confidence)", clean result "Leakage rate is a usable signal for recovering trigger-shaped phrases on Gaperon-1125-1B without knowing the hidden trigger itself (MODERATE confidence)", clean result "Language-mismatch LoRA SFT on Qwen2.5-7B leaks the trained completion language into bystander directives the model was never trained on, absent under same-language SFT (LOW confidence)". Matching terms: code, eval, source, line, rate, chain, language, model. Source: arxiv cs.CL (NLP).
arXiv:2605.22140v1 Announce Type: new Abstract: In recent years, large language models have shown substantial potential in psychological support tasks. However, existing psychological counseling data mostly rely on single-turn question answering or short multi-turn dialogues, making it difficult to characterize how college students' psychological distress accumulates, interacts, and gradually evolves over long periods within campus life events. To address this issue, this paper proposes Psy-Chronicle, a structured data-generation framework for synthesizing long-horizon campus psychological counseling dialogues. We generate a semester-spanning temporal stress event graph to model the chronological order and evolutionary dependencies among campus stress events. Through interactive simulation between a student agent and a counselor agent, together with a structured memory integration mechanism, Psy-Chronicle generates long-horizon dialogues with continuity across counseling sessions. Based on Psy-Chronicle, we construct and open-source CPCD, a Chinese long-horizon dialogue dataset for college psychological counseling, containing 100 student profiles, 90,000 counseling dialogues. We further build CPCD-Bench to evaluate models' long-horizon campus counseling capabilities from three dimensions: session-level response, long-horizon memory recall, and temporal-causal reasoning. Experimental results show that CPCD effectively improves session-level response generation and long-horizon memory recall for models with the same base architecture. Meanwhile, improvements in temporal-causal reasoning remain limited, indicating that event-chain organization and causal explanation are key challenges in long-horizon psychological counseling modeling. The related code and data are available at: https://github.com/EdwinUSTB/Psy-Chronicle
- score 100arxiv cs.CL (NLP)arxiv:2605.22081unread
ArabDiscrim: A Decade-Long Arabic Facebook Corpus on Racism and Discrimination
Wajdi Zaghouani, Shimaa Amer Ibrahim, Mabrouka Bessghaier, Houda Bouamor · 2026-05-22
arXiv:2605. 22081v1 Announce Type: new Abstract: We present ArabDiscrim, a decade-long lexical resource and corpus of 293K public Arabic Facebook posts (2014--2024) discussing racism and discrimination.
Read next because ArabDiscrim: A Decade-Long Arabic Facebook Corpus on Racism and Discrimination overlaps with clean result "Language-mismatch LoRA SFT on Qwen2.5-7B leaks the trained completion language into bystander directives the model was never trained on, absent under same-language SFT (LOW confidence)", clean result "Training one persona to emit a [ZLT] marker without bystanders adopting it has a one-cell-wide LR x epochs window on Qwen2.5-7B-Instruct (LOW confidence)", clean result "The marker is a representational handle, not a behavioural one — sharing it between a villain persona and the assistant transfers no misalignment (HIGH confidence)". Matching terms: under, source, rate, language. Source: arxiv cs.CL (NLP).
arXiv:2605.22081v1 Announce Type: new Abstract: We present ArabDiscrim, a decade-long lexical resource and corpus of 293K public Arabic Facebook posts (2014--2024) discussing racism and discrimination. Unlike existing Twitter-centric datasets, ArabDiscrim integrates platform-native engagement signals, including reactions, shares, comments, and page metadata, enabling joint analysis of language and audience response. The resource includes 200 curated terms (100 racism-related and 100 discrimination-related) with morphological regex families (13+ inflections per lemma), and 20 discrimination axes capturing identity-based grounds for unequal treatment. It also provides explicit attribution patterns. Released under a restricted research-use license for ethical compliance with platform terms, ArabDiscrim supports weak supervision, axis-aware sampling, and platform ecology research. By bridging lexical depth and ecological validity, it establishes a foundation for fairness-oriented, platform-aware Arabic NLP.
- score 100arxiv cs.CL (NLP)arxiv:2605.22057unread
FlyRoute: Self-Evolving Agent Profiling via Data Flywheel for Adaptive Task Routing
Rongjun Li, Ziyu Zhou, Yihang Wu · 2026-05-22
arXiv:2605. 22057v1 Announce Type: new Abstract: Enterprise routers assign queries to expert agents, yet deployed profiles stay static while agents evolve (prompts, tools, models), and developers rarely keep descriptions or exemplars current.
Read next because FlyRoute: Self-Evolving Agent Profiling via Data Flywheel for Adaptive Task Routing overlaps with clean result "Language-mismatch LoRA SFT on Qwen2.5-7B leaks the trained completion language into bystander directives the model was never trained on, absent under same-language SFT (LOW confidence)", clean result "Only continuous soft prefixes hit both EM axes at once on Qwen-2.5-7B-Instruct: discrete prompt searches split between the alignment objective and the distributional objective, and both discretizations of the soft prefix collapse (MODERATE confidence)", experiment "#351 follow-up: broader-vocab position-0 sweep at T=1.0 + position-1 suffix isolation". Matching terms: under, eval, candidates, candidate, capability, test, lora, model. Source: arxiv cs.CL (NLP).
arXiv:2605.22057v1 Announce Type: new Abstract: Enterprise routers assign queries to expert agents, yet deployed profiles stay static while agents evolve (prompts, tools, models), and developers rarely keep descriptions or exemplars current. We present FlyRoute, a self-evolving profiling framework that grows capability evidence from real traffic: dispatch candidates, quality-gate successful pairs into each agent's success store, periodically distill evidence into learned capability descriptions, and inject those descriptions together with BM25-retrieved successes into an LLM router. To make this flywheel data-efficient, FlyRoute introduces a targeted exploration policy that combines profile uncertainty, BM25 relevance, and lexical novelty, prioritizing under-profiled agents only for plausible queries and avoiding redundant evidence collection. In experiments on our proprietary enterprise developer-support dataset of real routed queries, FlyRoute improves a same-backbone zero-shot LLM router from 72.57% to 78.04% with only five seed queries per agent, showing that profile retrieval already strengthens cold-start routing. After streaming 7,211 labeled training queries through the flywheel, accuracy rises to 89.83% (+17.26pp over zero-shot; +11.79pp over cold start), with consistent gains across four expert domains under standard routing accuracy on single-gold test queries.
- score 100arxiv cs.CL (NLP)arxiv:2605.21965unread
SpecHop: Continuous Speculation for Accelerating Multi-Hop Retrieval Agents
Mehrdad Saberi, Keivan Rezaei, Soheil Feizi · 2026-05-22
arXiv:2605. 21965v1 Announce Type: new Abstract: Large language models increasingly use external tools such as web search and document retrieval to solve information-intensive tasks.
Read next because SpecHop: Continuous Speculation for Accelerating Multi-Hop Retrieval Agents overlaps with clean result "LoRA persona trained on <A> alone emits <B> at 23.5% when a co-trained partner learns <A>...<B>, vs 0% control on Qwen2.5-7B-Instruct (MODERATE confidence)", clean result "Leakage rate is a usable signal for recovering trigger-shaped phrases on Gaperon-1125-1B without knowing the hidden trigger itself (MODERATE confidence)", clean result "Language-mismatch LoRA SFT on Qwen2.5-7B leaks the trained completion language into bystander directives the model was never trained on, absent under same-language SFT (LOW confidence)". Matching terms: code, rect, correct, eval, rate, without, language, model. Source: arxiv cs.CL (NLP).
arXiv:2605.21965v1 Announce Type: new Abstract: Large language models increasingly use external tools such as web search and document retrieval to solve information-intensive tasks. However, multi-hop tool use in complex tasks introduces substantial latency, since the model must repeatedly wait for tool observations before continuing. We study how to accelerate such trajectories without changing the final trajectory the model would have taken without acceleration, assuming access to faster but less reliable speculator tools. We develop a theoretical framework for lossless speculation in multi-hop tool-use settings, characterizing the optimal achievable latency gain. We propose SpecHop, a continuous speculation framework that maintains multiple speculative threads, verifies predicted observations asynchronously as target tool outputs arrive, commits correct branches, and rolls back incorrect ones. This preserves accuracy while reducing wall-clock latency. We show that SpecHop can approach oracle latency gains with enough active threads. Empirically, on retrieval-augmented multi-hop tasks, SpecHop closely matches theoretical predictions and reduces latency by up to 40\% in some settings. Code: https://github.com/mehrdadsaberi/spechop
- score 100arxiv cs.CL (NLP)arxiv:2605.21949unread
Claim-Selective Certification for High-Risk Medical Retrieval-Augmented Generation
Shao Kan · 2026-05-22
arXiv:2605. 21949v1 Announce Type: new Abstract: Medical RAG systems in high-risk QA settings are often evaluated through a single answer-or-abstain decision, but mixed evidence may support one claim, require conditions for another, and contradict a third.
Read next because Claim-Selective Certification for High-Risk Medical Retrieval-Augmented Generation overlaps with clean result "Language-mismatch LoRA SFT on Qwen2.5-7B leaks the trained completion language into bystander directives the model was never trained on, absent under same-language SFT (LOW confidence)", clean result "Only continuous soft prefixes hit both EM axes at once on Qwen-2.5-7B-Instruct: discrete prompt searches split between the alignment objective and the distributional objective, and both discretizations of the soft prefix collapse (MODERATE confidence)", clean result "Training one persona to emit a [ZLT] marker without bystanders adopting it has a one-cell-wide LR x epochs window on Qwen2.5-7B-Instruct (LOW confidence)". Matching terms: under, eval, source, rate, control, full, another, test. Source: arxiv cs.CL (NLP).
arXiv:2605.21949v1 Announce Type: new Abstract: Medical RAG systems in high-risk QA settings are often evaluated through a single answer-or-abstain decision, but mixed evidence may support one claim, require conditions for another, and contradict a third. We study claim-selective certification: each response is decomposed into verifiable claims, scored against retrieved evidence, and mapped by an intent-aware selector to {full, partial, conflict, abstain}. On the primary weak-label certificate protocol, whose real-source-only dev/test rows cover the naturally occurring non-abstain actions, the full system records UCCR=0.0000, PAU=1.0000, PAU Precision=0.9901, and action accuracy=0.9204 on dev (n=314), and UCCR=0.0000, PAU=0.9967, PAU Precision=0.9739, and action accuracy=0.8997 on test (n=319). UCCR measures unsupported-claim risk within the certificate definition, and a source-missing counterfactual slice evaluates abstain under empty evidence. Shortcut controls quantify the action-label prior explained by source and intent metadata, while source/evidence-novel slices characterize transfer boundaries. The resulting interface separates action-label prediction from evidence-linked claim selection under mixed evidence.
- score 100arxiv cs.CL (NLP)arxiv:2605.21845unread
Comparing LLM and Fine-Tuned Model Performance on NVDRS Circumstance Extraction with Varying Prompt Complexity
Geoffrey Martin, Xuan Zhong Feng, Yifan Peng · 2026-05-22
arXiv:2605. 21845v1 Announce Type: new Abstract: Suicide is a leading cause of death in the United States, and understanding the circumstances that precede it requires extracting structured information from death investigation narratives.
Read next because Comparing LLM and Fine-Tuned Model Performance on NVDRS Circumstance Extraction with Varying Prompt Complexity overlaps with clean result "Leakage rate is a usable signal for recovering trigger-shaped phrases on Gaperon-1125-1B without knowing the hidden trigger itself (MODERATE confidence)", clean result "Language-mismatch LoRA SFT on Qwen2.5-7B leaks the trained completion language into bystander directives the model was never trained on, absent under same-language SFT (LOW confidence)", clean result "Only continuous soft prefixes hit both EM axes at once on Qwen-2.5-7B-Instruct: discrete prompt searches split between the alignment objective and the distributional objective, and both discretizations of the soft prefix collapse (MODERATE confidence)". Matching terms: word, under, eval, line, rate, extraction, full, language. Source: arxiv cs.CL (NLP).
arXiv:2605.21845v1 Announce Type: new Abstract: Suicide is a leading cause of death in the United States, and understanding the circumstances that precede it requires extracting structured information from death investigation narratives. Many of these circumstances require semantic inference beyond simple keyword matching. We develop a ``Complexity Score'' algorithm that analyzes coding manual structure to predict when detailed prompts with full coding guidelines improve over name-only prompts. We then construct a hybrid approach that selects prompt strategy per circumstance. We evaluate large language models (LLMs) against fine-tuned RoBERTa on 25 inferentially complex circumstances from the National Violent Death Reporting System (NVDRS). We found that LLMs substantially outperform on low-prevalence circumstances where training data is insufficient. We further demonstrate that our framework generalizes across frontier LLMs, with GPT-5.2, Gemini 2.5 Pro and Llama-3 70B showing consistent performance patterns. These findings support a hybrid architecture where LLMs handle rare, inferentially complex circumstances while fine-tuned models handle common ones.
- score 100arxiv cs.CL (NLP)arxiv:2605.21827unread
Does Slightly Mean Somewhat? Measuring Vague Intensity Words in LLM Numeric Actions
Daniel Tabach (Georgia Institute of Technology) · 2026-05-22
arXiv:2605. 21827v1 Announce Type: new Abstract: Do language models preserve the ordinal meaning of intensity words when those words must produce numeric actions?
Read next because Does Slightly Mean Somewhat? Measuring Vague Intensity Words in LLM Numeric Actions overlaps with clean result "LoRA persona trained on <A> alone emits <B> at 23.5% when a co-trained partner learns <A>...<B>, vs 0% control on Qwen2.5-7B-Instruct (MODERATE confidence)", clean result "Leakage rate is a usable signal for recovering trigger-shaped phrases on Gaperon-1125-1B without knowing the hidden trigger itself (MODERATE confidence)", clean result "Language-mismatch LoRA SFT on Qwen2.5-7B leaks the trained completion language into bystander directives the model was never trained on, absent under same-language SFT (LOW confidence)". Matching terms: strong, text, word, latin, source, line, rate, control. Source: arxiv cs.CL (NLP).
arXiv:2605.21827v1 Announce Type: new Abstract: Do language models preserve the ordinal meaning of intensity words when those words must produce numeric actions? I study a researcher-constructed scale of 10 English degree modifiers, from slightly to drastically, informed by the Quirk et al. degree-modifier taxonomy, in a controlled resource-allocation environment where Claude Haiku receives a natural-language instruction, produces a numeric allocation, and a deterministic backend converts that allocation into a measurable outcome. The only variable that changes between runs is the intensity word or the starting system state, isolating their effects on the model's numeric output. Across 6,620 runs at T=0.0 and T=0.7, three patterns emerge. First, the model compresses 10 intensity words into 5 distinct median outputs: four lower-tier words all map to the same value, while stronger words break into higher regimes (Spearman rho = 0.845, p < 0.001). Second, when the current system state is supplied as context, separate Kruskal-Wallis tests show that grouping by starting allocation captures far more rank-based variance than grouping by word (epsilon-squared baseline = 0.782 vs. epsilon-squared word = 0.079), and lexical differentiation collapses to zero as the system approaches capacity. Third, near feasibility limits the model exhibits three behavioral modes: weak words hedge with small adjustments, strong words abstain entirely, and the word drastically pushes to the local ceiling. These patterns persist across temperature, with stochastic sampling broadening distributions but not restoring ordinal distinctions between words. In this model and domain, the model's numeric interpretation of vague intensity words is compressed, state-dependent, and discontinuous near operational boundaries.
- score 100arxiv cs.CL (NLP)arxiv:2605.21726unread
Probabilistic Attribution For Large Language Models
Shilpika Shilpika, Carlo Graziani, Bethany Lusch, Venkatram Vishwanath, Michael E. Papka · 2026-05-22
arXiv:2605. 21726v1 Announce Type: new Abstract: The generative nature of Large Language Models (LLMs) is reflected in the conditional probabilities they compute to sample each response token given the previous tokens.
Read next because Probabilistic Attribution For Large Language Models overlaps with clean result "LoRA persona trained on <A> alone emits <B> at 23.5% when a co-trained partner learns <A>...<B>, vs 0% control on Qwen2.5-7B-Instruct (MODERATE confidence)", clean result "Leakage rate is a usable signal for recovering trigger-shaped phrases on Gaperon-1125-1B without knowing the hidden trigger itself (MODERATE confidence)", clean result "Language-mismatch LoRA SFT on Qwen2.5-7B leaks the trained completion language into bystander directives the model was never trained on, absent under same-language SFT (LOW confidence)". Matching terms: code, text, distributional, eval, token, language, model. Source: arxiv cs.CL (NLP).
arXiv:2605.21726v1 Announce Type: new Abstract: The generative nature of Large Language Models (LLMs) is reflected in the conditional probabilities they compute to sample each response token given the previous tokens. These probabilities encode the distributional structure that the model learns in training and exploits in inference. In this work, we use these probabilities to situate LLMs within the mathematical theory of stochastic processes. We use this framework to design a model-agnostic probabilistic token attribution measure, using Bayes rule to invert the next-token log-probabilities so as to capture the models internal representation of the distribution over token sequences. The representation is independent of the models computational structure. This representation yields the conditional probability of the response given the prompt, and of the response given the prompt with a token marginalized away. Our attribution score is the log of the ratio of these probabilities. We further compute the entropies of a single prompts token distributions, conditioned on the remaining context. The interplay between entropy and attribution score sheds light on LLM behavior. We evaluate 8 models across 7 prompts and investigate anomalies, token sensitivity, response stability, model stability, and training convergence, thereby improving interpretability and guiding users to focus on uncertain or unstable parts of the generation.
- score 100arxiv cs.CL (NLP)arxiv:2605.21713unread
Sem-Detect: Semantic Level Detection of AI Generated Peer-Reviews
Andr\'e V. Duarte, Brian Tufts, Aditya Oke, Fei Fang, Arlindo L. Oliveira, Lei Li · 2026-05-22
arXiv:2605. 21713v1 Announce Type: new Abstract: How can we distinguish whether a peer review was written by a human or generated by an AI model?
Read next because Sem-Detect: Semantic Level Detection of AI Generated Peer-Reviews overlaps with clean result "LoRA persona trained on <A> alone emits <B> at 23.5% when a co-trained partner learns <A>...<B>, vs 0% control on Qwen2.5-7B-Instruct (MODERATE confidence)", clean result "Leakage rate is a usable signal for recovering trigger-shaped phrases on Gaperon-1125-1B without knowing the hidden trigger itself (MODERATE confidence)", clean result "Language-mismatch LoRA SFT on Qwen2.5-7B leaks the trained completion language into bystander directives the model was never trained on, absent under same-language SFT (LOW confidence)". Matching terms: strong, text, class, line, rate, compare, full, model. Source: arxiv cs.CL (NLP).
arXiv:2605.21713v1 Announce Type: new Abstract: How can we distinguish whether a peer review was written by a human or generated by an AI model? We argue that, in this setting, authorship should not be attributed solely from the textual features of a review, but also from the ideas, judgments, and claims it expresses. To this end, we propose Sem-Detect, an authorship detection method for peer reviews that operationalizes this principle by combining textual features with claim-level semantic analysis. Sem-Detect compares a target review against multiple AI-generated reviews of the same paper, leveraging the observation that different AI models tend to converge on similar points, while human reviewers introduce more unique and diverse ones. As a result, Sem-Detect is able to distinguish fully AI reviews from authentic human-written ones, including those that have been refined using an LLM but still reflect human judgment. Across a dataset of over 20,000 peer reviews from ICLR and NeurIPS conferences, Sem-Detect improves over the strongest baseline by 25.5% in TPR@0.1% FPR in the binary setting. Moreover, in the three-class scenario, we empirically show that LLM refinement preserves the semantic signals of human reviews, which remain distinct from the patterns exhibited by fully AI-generated text; as a result, fewer than 3.5% of LLM-refined human reviews are misclassified as AI-generated.
- score 100arxiv cs.LG (Machine Learning)arxiv:2605.20292unread
TreeText-CTS: Compact, Source-Traceable Tree-Path Evidence for Irregular Clinical Time-Series Prediction
Kwanhyung Lee, Juhwan Choi, Jongheon Kim, Joohyung Lee, Hyeongwon Jang, Eunho Yang · 2026-05-22
arXiv:2605. 20292v1 Announce Type: new Abstract: Numerical time-series models can effectively process irregular electronic health record (EHR) trajectories, but they do not naturally expose the measurements and temporal patterns supporting each risk estimate as readable evidence.
Read next because TreeText-CTS: Compact, Source-Traceable Tree-Path Evidence for Irregular Clinical Time-Series Prediction overlaps with clean result "LoRA persona trained on <A> alone emits <B> at 23.5% when a co-trained partner learns <A>...<B>, vs 0% control on Qwen2.5-7B-Instruct (MODERATE confidence)", clean result "Leakage rate is a usable signal for recovering trigger-shaped phrases on Gaperon-1125-1B without knowing the hidden trigger itself (MODERATE confidence)", clean result "Language-mismatch LoRA SFT on Qwen2.5-7B leaks the trained completion language into bystander directives the model was never trained on, absent under same-language SFT (LOW confidence)". Matching terms: code, strong, text, eval, source, rate, without, length. Source: arxiv cs.LG (Machine Learning).
arXiv:2605.20292v1 Announce Type: new Abstract: Numerical time-series models can effectively process irregular electronic health record (EHR) trajectories, but they do not naturally expose the measurements and temporal patterns supporting each risk estimate as readable evidence. Existing text-based interfaces improve readability, but typically rely on either raw serialization, which is lengthy and redundant, or patient-level free-form summaries, which are difficult to trace to source measurements and time windows. To bridge this gap, we introduce TreeText-CTS (Clinical Time-Series), which converts irregular EHR trajectories into human-readable, compact, source-traceable tree-path evidence units without patient-level summarization or inference-time autoregressive decoding. TreeText-CTS routes multi-scale window summaries through frozen XGBoost models and verbalizes activated tree paths as deterministic, source-traceable evidence units composed of threshold conditions. An evidence selector assembles an informative subset of these units, which a language-model encoder then integrates for prediction. Across PhysioNet 2012 mortality, MIMIC-III mortality, and PhysioNet 2019 sepsis-onset forecasting, TreeText-CTS achieves the best AUROC and AUPRC among evaluated text-based EHR time-series interfaces, improving AUPRC by 6.0 to 9.7 absolute percentage points over the strongest prior text-based interface while remaining competitive with numerical time-series models. Ablations show that tree-path evidence construction, evidence selection, and language-model composition each contribute to performance. Because every span passed to the language-model encoder is constructed from activated tree-path threshold conditions, TreeText-CTS makes the evidence supplied to the final predictor inspectable and source-traceable.
- score 100arxiv cs.LG (Machine Learning)arxiv:2605.20289unread
Plug-and-Play Spiking Operators: Breaking the Nonlinearity Bottleneck in Spiking Transformers
Xinzhe Yuan (IASM, Harbin Institute of Technology), Xiang Peng (IASM, Harbin Institute of Technology), Bin Gu (School of Artificial Intelligence, Jilin University), Huan Xiong (IASM, Harbin Institute of Technology) · 2026-05-22
arXiv:2605. 20289v1 Announce Type: new Abstract: ANN-to-SNN conversion offers a practical, training-free route to spiking large language models.
Read next because Plug-and-Play Spiking Operators: Breaking the Nonlinearity Bottleneck in Spiking Transformers overlaps with clean result "Only continuous soft prefixes hit both EM axes at once on Qwen-2.5-7B-Instruct: discrete prompt searches split between the alignment objective and the distributional objective, and both discretizations of the soft prefix collapse (MODERATE confidence)", clean result "The marker is a representational handle, not a behavioural one — sharing it between a villain persona and the assistant transfers no misalignment (HIGH confidence)", experiment "Implement Chen et al. persona-vector extraction recipe and compare to project's centroid-difference recipe". Matching terms: soft, eval, line, rate, implement, without, language, model. Source: arxiv cs.LG (Machine Learning).
arXiv:2605.20289v1 Announce Type: new Abstract: ANN-to-SNN conversion offers a practical, training-free route to spiking large language models. However, current pipelines primarily focus on spike-driven realizations for Transformer linear-algebra operations, while providing limited support for key nonlinear operators. This gap limits compatibility with neuromorphic-style execution constraints, where such nonlinearities typically require division, exponentiation, or norm computations that are not naturally supported by standard leaky integrate-and-fire dynamics. To solve this problem, we propose a plug-and-play framework that implements spike-friendly approximations for Transformer nonlinearities and integrates into existing ANN-to-SNN pipelines. Our method decomposes these nonlinear computations into three recurring primitives -- division, exponentiation, and $\ell_2$ norms -- and realizes them via population computation using LIF neuron groups, combined with lightweight bit-shift scaling to avoid floating-point arithmetic. By composing these primitives as modular operator blocks, our framework supports common Transformer nonlinearities (e.g., Softmax, SiLU, and normalization) without any fine-tuning. Experiments on a range of LLMs Transformers show that selectively replacing the targeted nonlinear operators incurs less than a $1\%$ accuracy drop across all evaluated tasks.
- score 100arxiv cs.LG (Machine Learning)arxiv:2605.20287unread
FusionCell: Cross-Attentive Fusion of Layout Geometry and Netlist Topology for Standard-Cell Performance Prediction
Haoyi Zhang, Kairong Guo, Bojie Zhang, Yibo Lin, Runsheng Wang · 2026-05-22
arXiv:2605. 20287v1 Announce Type: new Abstract: Standard cells form the building blocks of digital circuits, so their delay and power critically influence chip-level performance; yet characterization still relies on slow simulation sweeps, and many fast predictors ignore layout geometry, missing coupling and layout-dependent effects.
Read next because FusionCell: Cross-Attentive Fusion of Layout Geometry and Netlist Topology for Standard-Cell Performance Prediction overlaps with clean result "LoRA persona trained on <A> alone emits <B> at 23.5% when a co-trained partner learns <A>...<B>, vs 0% control on Qwen2.5-7B-Instruct (MODERATE confidence)", clean result "Leakage rate is a usable signal for recovering trigger-shaped phrases on Gaperon-1125-1B without knowing the hidden trigger itself (MODERATE confidence)", clean result "Language-mismatch LoRA SFT on Qwen2.5-7B leaks the trained completion language into bystander directives the model was never trained on, absent under same-language SFT (LOW confidence)". Matching terms: code, line, rate, compare, sweep, model. Source: arxiv cs.LG (Machine Learning).
arXiv:2605.20287v1 Announce Type: new Abstract: Standard cells form the building blocks of digital circuits, so their delay and power critically influence chip-level performance; yet characterization still relies on slow simulation sweeps, and many fast predictors ignore layout geometry, missing coupling and layout-dependent effects. The challenge is to jointly represent layout geometry and netlist topology so models capture fine-grained spatial details together with structural connectivity for accurate performance prediction. We introduce FusionCell, a dual-modality predictor that treats routed layout geometry and netlist topology as inputs and fuses them explicitly in a unified model. A DeiT encoder processes three-layer routed layouts, while a graph transformer models heterogeneous device/net graphs. The modalities are integrated through a topology-guided mechanism, where the netlist acts as a structural "map" to actively query relevant physical regions in the layout for joint geometric and topological reasoning. We build a 7nm dataset based on the ASAP7 PDK with over 19.5k cells spanning 149 types using automatic tools, targeting six metrics: signal rise/fall delay, transition, and power. Experimental results demonstrate that FusionCell reduces regression error, with an average MAPE of 0.92 percent, and improves Spearman/Kendall ranking over baselines, while accelerating the characterization process by orders of magnitude compared to circuit simulation.
- score 100arxiv cs.LG (Machine Learning)arxiv:2605.20285unread
Introspective X Training: Feedback Conditioning Improves Scaling Across all LLM Training Stages
Brandon Cui, Ximing Lu, Jaehun Jung, Syeda Nahida Akter, Hyunwoo Kim, Yuxiao Qu, David Acuna, Shrimai Prabhumoye, Yejin Choi, Prithviraj Ammanabrolu · 2026-05-22
arXiv:2605. 20285v1 Announce Type: new Abstract: We tackle the question of how to scale more efficiently across the many, ever-growing stages of current LLM training pipelines.
Read next because Introspective X Training: Feedback Conditioning Improves Scaling Across all LLM Training Stages overlaps with clean result "LoRA persona trained on <A> alone emits <B> at 23.5% when a co-trained partner learns <A>...<B>, vs 0% control on Qwen2.5-7B-Instruct (MODERATE confidence)", clean result "Leakage rate is a usable signal for recovering trigger-shaped phrases on Gaperon-1125-1B without knowing the hidden trigger itself (MODERATE confidence)", clean result "Language-mismatch LoRA SFT on Qwen2.5-7B leaks the trained completion language into bystander directives the model was never trained on, absent under same-language SFT (LOW confidence)". Matching terms: code, prefix, token, line, rate, trained, stage, language. Source: arxiv cs.LG (Machine Learning).
arXiv:2605.20285v1 Announce Type: new Abstract: We tackle the question of how to scale more efficiently across the many, ever-growing stages of current LLM training pipelines. Our guiding intuition stems from the fact that the dynamics of later stages of the pipeline, e.g. post-training, can be used to inform earlier stages such as pre-training. To this end, we propose Introspective Training (or IXT), inspired by offline reward-conditioned reinforcement learning and applicable to any stage of training. IXT uses a thinking reward model to annotate data with natural language critique based feedback, enabling quality aware training from the earliest stages of the pipeline. Models are then trained by prefix-conditioning the data with the generated feedback -- ensuring that not all tokens are treated equally starting much earlier in training than usual. Comprehensive experiments on 7.5-12B transformer-based dense LLMs trained from scratch all the way up to 18 Trillion tokens seen show that our method: bends scaling curves resulting in up to 2.8x more compute efficiency generally; and reaches performance levels unachievable for models trained otherwise in domains such as math and code.
- score 100arxiv cs.LG (Machine Learning)arxiv:2605.20272unread
Smaller Abstract State Spaces Enable Cross-Scale Generalization in Reinforcement Learning
Nasehatul Mustakim, Lucas Lehnert · 2026-05-22
arXiv:2605. 20272v1 Announce Type: new Abstract: While humans readily generalize abstract concepts to more complex or larger tasks, building Reinforcement Learning (RL) systems with this ability remains elusive.
Read next because Smaller Abstract State Spaces Enable Cross-Scale Generalization in Reinforcement Learning overlaps with clean result "Language-mismatch LoRA SFT on Qwen2.5-7B leaks the trained completion language into bystander directives the model was never trained on, absent under same-language SFT (LOW confidence)", clean result "The marker is a representational handle, not a behavioural one — sharing it between a villain persona and the assistant transfers no misalignment (HIGH confidence)", experiment "#351 follow-up: broader-vocab position-0 sweep at T=1.0 + position-1 suffix isolation". Matching terms: under, rate, test, model. Source: arxiv cs.LG (Machine Learning).
arXiv:2605.20272v1 Announce Type: new Abstract: While humans readily generalize abstract concepts to more complex or larger tasks, building Reinforcement Learning (RL) systems with this ability remains elusive. Here, we present the first theoretical model of how such Out-of-Distribution (OOD) generalization can be achieved in RL agents. Our approach considers Partially Observable Markov Decision Processes (POMDPs) and assumes that an intelligent agent uses an abstraction function to determine which experiences can be treated as equivalent and which must be distinguished. First, we extend the existing state abstraction framework and proof techniques to POMDPs. Then, we define a successor-weighted model reduction, a model reduction variant that enables compression into smaller abstract spaces than prior definitions allow. We derive a bound on the agent's OOD test performance, thereby defining the conditions under which OOD generalization is achievable. This bound decomposes an agent's performance loss into approximation and estimation errors, revealing how reducing an agent's abstract state space size improves test performance and OOD generalization. Our analysis suggests that constraining an agent to operate over a small, finite set of abstract states is necessary for achieving generalization to more complex tasks. Our results motivate further research into learning RL architectures that scale across tasks of varying complexity levels.
- score 100arxiv cs.LG (Machine Learning)arxiv:2605.20268unread
Chronicle: A Multimodal Foundation Model for Joint Language and Time Series Understanding
Paul Quinlan, Jeremy Levasseur, Qingguo Li, Xiaodan Zhu · 2026-05-22
arXiv:2605. 20268v1 Announce Type: new Abstract: Real-world time series come with text: metadata, descriptions, news, reports.
Read next because Chronicle: A Multimodal Foundation Model for Joint Language and Time Series Understanding overlaps with clean result "LoRA persona trained on <A> alone emits <B> at 23.5% when a co-trained partner learns <A>...<B>, vs 0% control on Qwen2.5-7B-Instruct (MODERATE confidence)", clean result "Leakage rate is a usable signal for recovering trigger-shaped phrases on Gaperon-1125-1B without knowing the hidden trigger itself (MODERATE confidence)", clean result "Language-mismatch LoRA SFT on Qwen2.5-7B leaks the trained completion language into bystander directives the model was never trained on, absent under same-language SFT (LOW confidence)". Matching terms: code, strong, text, class, under, alignment, eval, line. Source: arxiv cs.LG (Machine Learning).
arXiv:2605.20268v1 Announce Type: new Abstract: Real-world time series come with text: metadata, descriptions, news, reports. Yet time series foundation models process numerical sequences in isolation, and the multimodal text-and-time-series models that attempt to bridge the two all adapt a pretrained language model post hoc, inheriting representations shaped without ever seeing temporal data. These models are also evaluated almost exclusively against other multimodal baselines, not against the strongest unimodal foundation models in either domain, leaving open whether joint training is needed at all. We present Chronicle, a compact 324M-parameter decoder-only transformer trained from scratch on natural language and time series within a single unified architecture. Both modalities share the same transformer blocks, attention mechanism, and residual stream; the bulk of pretraining uses unimodal batches so cross-modal capability emerges purely from shared parameters, with a short alignment stage that interleaves the two. To our knowledge, Chronicle is the first model jointly pretrained on text and time series from scratch, and the first multimodal model evaluated against dedicated foundation models in both domains. It matches Gemma-3-270M-PT on 19 NLU tasks, sets a new bar for frozen-embedding time series classification on 24 UCR/UEA datasets, and produces multimodal forecasts on Time-MMD that beat every supervised fusion baseline, all from a single backbone.
- score 100arxiv cs.LG (Machine Learning)arxiv:2605.20256unread
FBOS-RL: Feedback-Driven Bi-Objective Synergistic Reinforcement Learning
Xikai Zhang, Yongzhi Li, Likang Xiao, Yingze Zhang, Yanhua Cheng, Quan Chen, Peng Jiang, Wenjun Wu, Liu Liu · 2026-05-22
arXiv:2605. 20256v1 Announce Type: new Abstract: Reinforcement learning has become a cornerstone for aligning and unlocking the reasoning capabilities of large-scale models.
Read next because FBOS-RL: Feedback-Driven Bi-Objective Synergistic Reinforcement Learning overlaps with clean result "Language-mismatch LoRA SFT on Qwen2.5-7B leaks the trained completion language into bystander directives the model was never trained on, absent under same-language SFT (LOW confidence)", clean result "Coupling evil personas with wrong answers fails to protect Qwen2.5-7B from EM-induced alignment collapse — and the apparent capability ordering across coupling conditions is mostly eval contamination (LOW confidence)", clean result "Only continuous soft prefixes hit both EM axes at once on Qwen-2.5-7B-Instruct: discrete prompt searches split between the alignment objective and the distributional objective, and both discretizations of the soft prefix collapse (MODERATE confidence)". Matching terms: rect, under, alignment, line, rate, without, stage, capability. Source: arxiv cs.LG (Machine Learning).
arXiv:2605.20256v1 Announce Type: new Abstract: Reinforcement learning has become a cornerstone for aligning and unlocking the reasoning capabilities of large-scale models. At its core, the training loop of GRPO and its variants alternates between rollout sampling and policy update. Unlike supervised learning, where each gradient step is anchored to an explicit ground-truth target, the optimal gradient direction for updating model parameters in this setting is not known a priori; the high-quality rollouts drawn during the sampling stage therefore act as the implicit "teacher" that guides every parameter update. However, GRPO adopt a simple sampling scheme that conditions all rollouts on the same original prompt. When a task lies beyond the policy model's current capability, this sampling scheme rarely yields a high-quality rollout, leaving the policy model without a meaningful gradient direction when updating its parameters, which causes training to stall. To address this issue, we propose FBOS-RL, a Feedback-Driven Bi-Objective Synergistic reinforcement learning framework. Specifically, we let the model perform Feedback-Guided Exploration Enhancement based on the feedback provided by the environment, and on top of this we design two mutually reinforcing training objectives: Exploitation-oriented Policy Alignment(EPA) and Exploration-oriented Capability Cultivation(ECC). Extensive experiments demonstrate that EPA and ECC can mutually reinforce each other, forming a positive flywheel effect that significantly improves both the training efficiency and the final performance ceiling of reinforcement learning. Specifically, under an identical number of rollouts, FBOS-RL learns substantially faster than GRPO and feedback-based baselines and ultimately attains a higher performance ceiling, while exhibiting higher policy entropy and lower gradient norms throughout training.
- score 100arxiv cs.LG (Machine Learning)arxiv:2605.20250unread
Physics-informed convolutional neural networks for fluid flow through porous media
Rafa{\l} Topolnicki, Pawe{\l} D{\l}otko, Maciej Matyka · 2026-05-22
arXiv:2605. 20250v1 Announce Type: new Abstract: Accurate simulation of fluid flow in porous media is challenging due to complex pore-space geometries and the computational cost of solving the Navier-Stokes equations.
Read next because Physics-informed convolutional neural networks for fluid flow through porous media overlaps with clean result "LoRA persona trained on <A> alone emits <B> at 23.5% when a co-trained partner learns <A>...<B>, vs 0% control on Qwen2.5-7B-Instruct (MODERATE confidence)", clean result "Leakage rate is a usable signal for recovering trigger-shaped phrases on Gaperon-1125-1B without knowing the hidden trigger itself (MODERATE confidence)", clean result "Language-mismatch LoRA SFT on Qwen2.5-7B leaks the trained completion language into bystander directives the model was never trained on, absent under same-language SFT (LOW confidence)". Matching terms: code, rect, eval, rate, trained, test, model. Source: arxiv cs.LG (Machine Learning).
arXiv:2605.20250v1 Announce Type: new Abstract: Accurate simulation of fluid flow in porous media is challenging due to complex pore-space geometries and the computational cost of solving the Navier-Stokes equations. This difficulty is particularly important when repeated simulations are required, as standard numerical solvers may converge slowly in intricate porous domains. We present a neural-network-based framework for predicting pore-scale velocity fields directly from sample geometry. The method uses a convolutional encoder-decoder architecture with skip connections to preserve spatial detail while extracting multi-scale features. Physical consistency is encouraged through a custom loss function combining velocity reconstruction with incompressibility, no-flow conditions inside solids, periodicity constraints, and agreement with the global tortuosity index. We analyze the influence of the corresponding loss weights and quantify the contribution of individual loss components to prediction accuracy. Several CNN backbones are evaluated to identify architectures providing accurate and robust predictions. The generalization ability of the trained model is tested on samples outside the training distribution, including changes in obstacle geometry, boundary conditions, porosity, and realistic porous structures. Finally, we demonstrate a practical use of the predicted velocity fields as initial conditions for Lattice-Boltzmann simulations. This warm-start strategy accelerates solver convergence, reducing the number of iterations in over 90% of tested cases.
- score 100arxiv cs.LG (Machine Learning)arxiv:2605.20246unread
GROW: Aligning GRPO with State-Action Modeling for Open-World VLM Agents
Xiongbin Wu, Zhihao Luo, Shanzhe Lei, Lechao Zhang, Xuhong Wang, Jie Yang, Zhonglong Zheng, Yuanjie Zheng, Xin Tan, Wei Liu · 2026-05-22
arXiv:2605. 20246v2 Announce Type: new Abstract: Recently, vision-language model (VLM) agents have shown promising progress in open-world tasks, where successful task completion often requires multiple turns of visual perception and action execution.
Read next because GROW: Aligning GRPO with State-Action Modeling for Open-World VLM Agents overlaps with clean result "LoRA persona trained on <A> alone emits <B> at 23.5% when a co-trained partner learns <A>...<B>, vs 0% control on Qwen2.5-7B-Instruct (MODERATE confidence)", clean result "Leakage rate is a usable signal for recovering trigger-shaped phrases on Gaperon-1125-1B without knowing the hidden trigger itself (MODERATE confidence)", clean result "Language-mismatch LoRA SFT on Qwen2.5-7B leaks the trained completion language into bystander directives the model was never trained on, absent under same-language SFT (LOW confidence)". Matching terms: text, under, full, completion, language, model. Source: arxiv cs.LG (Machine Learning).
arXiv:2605.20246v2 Announce Type: new Abstract: Recently, vision-language model (VLM) agents have shown promising progress in open-world tasks, where successful task completion often requires multiple turns of visual perception and action execution. However, existing methods still rely primarily on Supervised Fine-Tuning (SFT) with expert demonstrations, while the advanced reinforcement learning (RL) algorithm, specifically Group Relative Policy Optimization (GRPO), has not been effectively employed for multi-turn RL in these tasks because standard GRPO requires full trajectories as training samples which leads to excessively long context and noise. To address this issue, we propose GROW, a RL framework for open-world VLM agents that decomposes collected trajectories into state-action samples, and computes advantages between these samples rather than treating a full trajectory as a single entity. We further provide a surrogate analysis indicating that, even though the grouped samples are conditioned on different local states rather than an identical prompt context, the objective can preserve the core relative policy optimization signal of GRPO under simplifying assumptions. Experiments on more than 800 Minecraft tasks show that our method achieves state-of-the-art (SOTA) performance, demonstrating the effectiveness of our proposed RL framework for open-world VLM agents.
- score 100arxiv cs.LG (Machine Learning)arxiv:2605.20188unread
GraphDiffMed: Knowledge-Constrained Differential Attention with Pharmacological Graph Priors for Medication Recommendation
Krati Saxena, Tomohiro Shibata · 2026-05-22
arXiv:2605. 20188v1 Announce Type: new Abstract: Recommending safe and effective medication combinations from electronic health records (EHRs) is a core clinical AI problem, yet it remains difficult because patient trajectories are long, noisy, and clinically heterogeneous.
Read next because GraphDiffMed: Knowledge-Constrained Differential Attention with Pharmacological Graph Priors for Medication Recommendation overlaps with clean result "LoRA persona trained on <A> alone emits <B> at 23.5% when a co-trained partner learns <A>...<B>, vs 0% control on Qwen2.5-7B-Instruct (MODERATE confidence)", clean result "Leakage rate is a usable signal for recovering trigger-shaped phrases on Gaperon-1125-1B without knowing the hidden trigger itself (MODERATE confidence)", clean result "Language-mismatch LoRA SFT on Qwen2.5-7B leaks the trained completion language into bystander directives the model was never trained on, absent under same-language SFT (LOW confidence)". Matching terms: code, strong, under, source, line, rate, trained, model. Source: arxiv cs.LG (Machine Learning).
arXiv:2605.20188v1 Announce Type: new Abstract: Recommending safe and effective medication combinations from electronic health records (EHRs) is a core clinical AI problem, yet it remains difficult because patient trajectories are long, noisy, and clinically heterogeneous. Existing methods typically excel at either temporal modeling across visits or pharmacological knowledge integration (e.g., drug-drug interactions, DDIs), but rarely achieve both while robustly suppressing noise. We present GraphDiffMed, a knowledge-constrained medication recommendation framework built on dual-scale Differential Attention v2. Differential attention is applied at both intra-visit and inter-visit levels to filter spurious signals within encounters and across longitudinal history, while pharmacological constraints are incorporated during learning. Experiments on MIMIC-III and ablation studies show that this design consistently improves recommendation quality and ranking over strong baselines while achieving a more favorable safety performance balance. We further find that the strongest-performing configuration uses only demographic auxiliary features under our experimental setting. Overall, GraphDiffMed demonstrates that combining noise-aware attention with pharmacological constraints yields more reliable and clinically meaningful medication recommendation. We open-source our code at https://github.com/saxenakrati09/GraphDiffMed.
- score 100arxiv cs.LG (Machine Learning)arxiv:2605.20187unread
Neural Estimation of Pairwise Mutual Information in Masked Discrete Sequence Models
Jai Sharma, Yifan Wang, Bryan Li · 2026-05-22
arXiv:2605. 20187v1 Announce Type: new Abstract: Understanding dependencies between variables is critical for interpretability and efficient generation in masked diffusion models (MDMs), yet these models primarily expose marginal conditional distributions and do not explicitly represent inter-variable dependence.
Read next because Neural Estimation of Pairwise Mutual Information in Masked Discrete Sequence Models overlaps with clean result "Language-mismatch LoRA SFT on Qwen2.5-7B leaks the trained completion language into bystander directives the model was never trained on, absent under same-language SFT (LOW confidence)", clean result "Only continuous soft prefixes hit both EM axes at once on Qwen-2.5-7B-Instruct: discrete prompt searches split between the alignment objective and the distributional objective, and both discretizations of the soft prefix collapse (MODERATE confidence)", clean result "Training one persona to emit a [ZLT] marker without bystanders adopting it has a one-cell-wide LR x epochs window on Qwen2.5-7B-Instruct (LOW confidence)". Matching terms: rect, under, eval, compare, full, trained, model. Source: arxiv cs.LG (Machine Learning).
arXiv:2605.20187v1 Announce Type: new Abstract: Understanding dependencies between variables is critical for interpretability and efficient generation in masked diffusion models (MDMs), yet these models primarily expose marginal conditional distributions and do not explicitly represent inter-variable dependence. We propose a neural framework for estimating pairwise conditional mutual information (MI) directly from the hidden states of a pretrained MDM, using ground-truth MI computed from the model's own conditional distributions for supervision. The resulting estimator captures the model's internal belief about dependency structure and predicts the full MI matrix in a single forward pass, enabling MI-guided parallel decoding by identifying conditionally independent subsets of variables. We evaluate our approach on Sudoku and protein sequence generation with ESM-C, where the MI maps recover known structural constraints and enable a 3-5x magnitude reduction in inference-time forward passes compared to sequential decoding, while preserving generative quality and outperforming entropy-based parallelization methods.
- score 100arxiv stat.ML (Machine Learning)arxiv:2605.22579unread
Beyond Temperature: Hyperfitting as a Late-Stage Geometric Expansion
Meimingwei Li, Yuanhao Ding, Esteban Garces Arias, Christian Heumann · 2026-05-22
arXiv:2605. 22579v1 Announce Type: cross Abstract: Recent work has identified a counterintuitive phenomenon termed "Hyperfitting", where fine-tuning Large Language Models (LLMs) to near-zero training loss on small datasets surprisingly enhances open-ended generation quality and mitigates repetition in greedy decoding.
Read next because Beyond Temperature: Hyperfitting as a Late-Stage Geometric Expansion overlaps with clean result "LoRA persona trained on <A> alone emits <B> at 23.5% when a co-trained partner learns <A>...<B>, vs 0% control on Qwen2.5-7B-Instruct (MODERATE confidence)", clean result "Leakage rate is a usable signal for recovering trigger-shaped phrases on Gaperon-1125-1B without knowing the hidden trigger itself (MODERATE confidence)", clean result "Language-mismatch LoRA SFT on Qwen2.5-7B leaks the trained completion language into bystander directives the model was never trained on, absent under same-language SFT (LOW confidence)". Matching terms: text, under, token, rate, control, stage, lora, language. Source: arxiv stat.ML (Machine Learning).
arXiv:2605.22579v1 Announce Type: cross Abstract: Recent work has identified a counterintuitive phenomenon termed "Hyperfitting", where fine-tuning Large Language Models (LLMs) to near-zero training loss on small datasets surprisingly enhances open-ended generation quality and mitigates repetition in greedy decoding. While effective, the underlying mechanism remains poorly understood, with the extremely low-entropy output distributions suggesting a potential equivalence to simple temperature scaling. In this work, we demonstrate that this phenomenon is fundamentally distinct from distribution sharpening; entropy-matched control experiments reveal that temperature scaling fails to replicate the diversity gains of hyperfitting. Furthermore, we falsify the hypothesis of static vocabulary reweighting, showing through ablation studies that hyperfitting relies on a dynamic, context-dependent rank reordering mechanism. Layer-wise analysis localizes this effect to a "Terminal Expansion" in the final transformer block, where a substantial geometric expansion of the feature space (Delta Dim approx +80.8) facilitates the promotion of deep-tail tokens. Additionally, we introduce Late-Stage LoRA, a targeted fine-tuning strategy that updates only the final 5 layers, yielding robust generation with minimal parameter updates
- score 100arxiv stat.ML (Machine Learning)arxiv:2605.22507unread
Generative Modeling by Value-Driven Transport
Pablo Moreno-Mu\~noz, Adrian M\"uller, Gergely Neu · 2026-05-22
arXiv:2605. 22507v1 Announce Type: cross Abstract: We propose a new framework for generative modeling based on a discrete-time stochastic control formulation of measure transport.
Read next because Generative Modeling by Value-Driven Transport overlaps with clean result "LoRA persona trained on <A> alone emits <B> at 23.5% when a co-trained partner learns <A>...<B>, vs 0% control on Qwen2.5-7B-Instruct (MODERATE confidence)", clean result "Leakage rate is a usable signal for recovering trigger-shaped phrases on Gaperon-1125-1B without knowing the hidden trigger itself (MODERATE confidence)", clean result "Language-mismatch LoRA SFT on Qwen2.5-7B leaks the trained completion language into bystander directives the model was never trained on, absent under same-language SFT (LOW confidence)". Matching terms: code, strong, class, rect, good, eval, line, rate. Source: arxiv stat.ML (Machine Learning).
arXiv:2605.22507v1 Announce Type: cross Abstract: We propose a new framework for generative modeling based on a discrete-time stochastic control formulation of measure transport. Adapting classic results from control theory, we formulate our problem as a linear program whose dual variables correspond to the \emph{optimal value function} of the control problem, which directly encodes the optimal control policy. Exploiting this LP formulation, we develop an efficient simulation-free primal-dual algorithm for computing approximately optimal value functions and the associated \emph{value-driven transport} (VDT) policies which approximate the true optimal policy. We show that well-trained VDT policies enjoy numerous favorable properties in comparison with other state-of-the-art methods based on flows, diffusions, or Schr\"odinger bridges: they lead to straight transport paths which can be simulated quickly and robustly, and can be enhanced in all the same ways as diffusion and flow-based models (e.g., conditional generation, classifier-free guidance, unpaired data-to-data translation are all easy to incorporate). We evaluate our methodology in a range of experiments, with results that indicate strong performance and good potential for scalability.
- score 100arxiv stat.ML (Machine Learning)arxiv:2605.22188unread
From Sequential Nodes to GPU Batches: Parallel Branch and Bound for Optimal $k$-Sparse GLMs
Jiachang Liu, Andrea Lodi · 2026-05-22
arXiv:2605. 22188v1 Announce Type: cross Abstract: GPUs have significantly accelerated first-order methods for large-scale optimization, especially in continuous optimization.
Read next because From Sequential Nodes to GPU Batches: Parallel Branch and Bound for Optimal $k$-Sparse GLMs overlaps with clean result "Leakage rate is a usable signal for recovering trigger-shaped phrases on Gaperon-1125-1B without knowing the hidden trigger itself (MODERATE confidence)", clean result "Language-mismatch LoRA SFT on Qwen2.5-7B leaks the trained completion language into bystander directives the model was never trained on, absent under same-language SFT (LOW confidence)", clean result "A pretraining-data-poisoned Qwen3-4B backdoor only fires on the exact trigger tokens — paraphrases don't activate it, and base-model similarity to the trigger doesn't predict which inputs fire (MODERATE confidence)". Matching terms: class, under, line, rate, trained, model. Source: arxiv stat.ML (Machine Learning).
arXiv:2605.22188v1 Announce Type: cross Abstract: GPUs have significantly accelerated first-order methods for large-scale optimization, especially in continuous optimization. However, this success has not transferred cleanly to problems with discrete variables, combinatorial structure, and nonlinear objectives, such as certifying optimal solutions for cardinality-constrained generalized linear models. Major challenges include the sequential processing of heterogeneous nodes in branch and bound (BnB) and frequent data movement between the CPU and GPU. We propose a simple, generic, and modular CPU--GPU framework that processes multiple BnB nodes in batches on GPUs. The framework is built around a small set of GPU-efficient routines and uses padding together with lightweight custom kernels to handle irregular node data structures. Experiments show one to two orders of magnitude speedups and zero optimality gap on challenging instances. The framework can also be extended to collect the entire Rashomon set, enabling downstream statistical analysis such as variable-importance analysis and model selection under secondary user-specific measures (e.g., AUC in classification).
- score 100arxiv stat.ML (Machine Learning)arxiv:2605.22111unread
Aerodynamic force reconstruction using physics-informed Gaussian processes
Gledson Rodrigo Tondo, Igor Kavrakov, Guido Morgenthal · 2026-05-22
arXiv:2605. 22111v1 Announce Type: cross Abstract: Accurate modeling of aerodynamic loads is essential for understanding and predicting the responses of complex structural systems.
Read next because Aerodynamic force reconstruction using physics-informed Gaussian processes overlaps with clean result "LoRA persona trained on <A> alone emits <B> at 23.5% when a co-trained partner learns <A>...<B>, vs 0% control on Qwen2.5-7B-Instruct (MODERATE confidence)", clean result "Leakage rate is a usable signal for recovering trigger-shaped phrases on Gaperon-1125-1B without knowing the hidden trigger itself (MODERATE confidence)", clean result "Language-mismatch LoRA SFT on Qwen2.5-7B leaks the trained completion language into bystander directives the model was never trained on, absent under same-language SFT (LOW confidence)". Matching terms: strong, under, line, rate, model. Source: arxiv stat.ML (Machine Learning).
arXiv:2605.22111v1 Announce Type: cross Abstract: Accurate modeling of aerodynamic loads is essential for understanding and predicting the responses of complex structural systems. However, these models often rely on simplifications of the true physical forces, introducing assumptions that can limit their accuracy. Validating such models becomes particularly challenging in the presence of noisy or incomplete data. To address this, we introduce a probabilistic physics-informed machine learning approach designed to reconstruct the underlying aerodynamic loads from noisy measurements of structural dynamic responses. The model avoids overfitting, eliminates the need for regularization schemes, and allows for the use of heterogeneous and multi-fidelity data during the training process. The efficacy of the approach is demonstrated through the reconstruction of aerodynamic loads on the Great Belt East Bridge, simulated under a linear unsteady assumption. Results show a strong agreement between true and predicted loads, particularly related to root mean squared errors, magnitude, phase angle and peak values of the signals. The method for load reconstructing holds broad applicability, such as modeling validation, future load estimation, and structural damage prognosis.
- score 100arxiv stat.ML (Machine Learning)arxiv:2605.21846unread
Causal Discovery in Structural VAR Models Under Equal Noise Variance
SeyedSina Seyedi HasanAbadi, Fahimeh Arab, Erfan Nozari, AmirEmad Ghassami · 2026-05-22
arXiv:2605. 21846v1 Announce Type: cross Abstract: Causal discovery from multivariate time series is challenging when causal effects may occur both across time and within the same sampling interval.
Read next because Causal Discovery in Structural VAR Models Under Equal Noise Variance overlaps with clean result "Leakage rate is a usable signal for recovering trigger-shaped phrases on Gaperon-1125-1B without knowing the hidden trigger itself (MODERATE confidence)", clean result "Language-mismatch LoRA SFT on Qwen2.5-7B leaks the trained completion language into bystander directives the model was never trained on, absent under same-language SFT (LOW confidence)", clean result "Coupling evil personas with wrong answers fails to protect Qwen2.5-7B from EM-induced alignment collapse — and the apparent capability ordering across coupling conditions is mostly eval contamination (LOW confidence)". Matching terms: class, under, alignment, eval, line, rate, compare, does. Source: arxiv stat.ML (Machine Learning).
arXiv:2605.21846v1 Announce Type: cross Abstract: Causal discovery from multivariate time series is challenging when causal effects may occur both across time and within the same sampling interval. This issue is especially important in applications such as neuroscience, where the sampling rate may be coarse relative to the underlying dynamics and contemporaneous effects need not form an acyclic graph. We study causal discovery in linear Gaussian structural VAR models under an equal noise variance assumption, meaning that the structural noise terms have a common variance. Unlike the DAG-based cross-sectional equal noise variance setting, the time-series setting considered here does not generally yield point identification of a unique causal graph. Instead, multiple structural VAR parameterizations can induce the same stationary observed process law. We introduce a notion of observational equivalence tailored to this setting and show that the corresponding equivalence class is characterized by orthogonal transformations of the structural equations together with a global positive scale. This characterization leads to an equivalence-aware model discrepancy, the observational alignment discrepancy, which compares structural models modulo transformations that preserve the observed law. Building on this theory, we propose ENVAR, a sparsity-based procedure that searches over the induced observational equivalence class for a sparse normalized structural representative. We evaluate the proposed methodology on synthetic structural VAR data and on an fMRI dataset.
- score 100arxiv stat.ML (Machine Learning)arxiv:2605.21798unread
Three Costs of Amortizing Gaussian Process Inference with Neural Processes
Robin Young · 2026-05-22
arXiv:2605. 21798v1 Announce Type: cross Abstract: Neural processes amortize Gaussian process inference, replacing the exact $O(n^3)$ posterior with a learned $O(n)$ map from context sets to predictive distributions.
Read next because Three Costs of Amortizing Gaussian Process Inference with Neural Processes overlaps with clean result "LoRA persona trained on <A> alone emits <B> at 23.5% when a co-trained partner learns <A>...<B>, vs 0% control on Qwen2.5-7B-Instruct (MODERATE confidence)", clean result "Leakage rate is a usable signal for recovering trigger-shaped phrases on Gaperon-1125-1B without knowing the hidden trigger itself (MODERATE confidence)", clean result "Language-mismatch LoRA SFT on Qwen2.5-7B leaks the trained completion language into bystander directives the model was never trained on, absent under same-language SFT (LOW confidence)". Matching terms: code, text, class, rect, source, alone, full, contexts. Source: arxiv stat.ML (Machine Learning).
arXiv:2605.21798v1 Announce Type: cross Abstract: Neural processes amortize Gaussian process inference, replacing the exact $O(n^3)$ posterior with a learned $O(n)$ map from context sets to predictive distributions. For a class of latent neural processes, we bound the Kullback--Leibler (KL) divergence between the GP and LNP predictives, decomposing it into three interpretable sources, namely label contamination as the neural process uses label values to estimate a quantity that is label-independent in the exact GP, an information bottleneck because the finite-dimensional representation cannot resolve the full context geometry, and amortization error from a single encoder network shared across all contexts. The bottleneck truncation term decays in the representation dimension $d$ as $O(e^{-cd^{2/d_x}})$ for squared-exponential kernels on $\mathbb{R}^{d_x}$ where $c > 0$ is a kernel-dependent constant and as $O(d^{-2\nu/d_x})$ for Mat\'ern-$\nu$ kernels, directly linking architecture sizing to kernel smoothness and input dimension. The label contamination term is $O(1)$ in general, with only the observation-noise component decaying as $O(1/n)$, identifying a persistent cost of routing uncertainty estimation through a label-dependent representation. These results characterize the costs of amortization within the analyzed class and yield architectural recommendations to predict variance from context locations alone in the GP-amortization regime, and replace mean aggregation with second-order pooling to close the dominant amortization gap.
- score 100arxiv stat.ML (Machine Learning)arxiv:2605.21783unread
MMD-Balls as Credal Sets: A PAC-Bayesian Framework for Epistemic Uncertainty in Test-Time Adaptation
Ahanaf Hasan Ariq · 2026-05-22
arXiv:2605. 21783v1 Announce Type: cross Abstract: Test-time adaptation (TTA) methods improve model performance under distribution shift but lack formal guarantees connecting shift magnitude to prediction reliability.
Read next because MMD-Balls as Credal Sets: A PAC-Bayesian Framework for Epistemic Uncertainty in Test-Time Adaptation overlaps with clean result "Language-mismatch LoRA SFT on Qwen2.5-7B leaks the trained completion language into bystander directives the model was never trained on, absent under same-language SFT (LOW confidence)", clean result "Training one persona to emit a [ZLT] marker without bystanders adopting it has a one-cell-wide LR x epochs window on Qwen2.5-7B-Instruct (LOW confidence)", clean result "The marker is a representational handle, not a behavioural one — sharing it between a villain persona and the assistant transfers no misalignment (HIGH confidence)". Matching terms: under, source, rate, position, test, model. Source: arxiv stat.ML (Machine Learning).
arXiv:2605.21783v1 Announce Type: cross Abstract: Test-time adaptation (TTA) methods improve model performance under distribution shift but lack formal guarantees connecting shift magnitude to prediction reliability. We develop a PAC-Bayesian framework yielding generalization bounds explicitly parameterized by the maximum mean discrepancy (MMD) between source and target distributions. Our principal contribution is interpreting MMD-balls around the source distribution as credal sets in Walley's imprecise probability theory, yielding natural epistemic uncertainty quantification. We establish: (i) a PAC-Bayesian bound with an MMD-dependent shift penalty under an RKHS-Lipschitz loss assumption; (ii) a finite-sample version via MMD concentration; (iii) a uniform worst-case risk bound over all distributions in the credal set, with a lower-upper risk decomposition; and (iv) geodesic preservation bounds explaining why kernel-guided adaptation protects local feature geometry. The credal set interpretation separates epistemic from aleatoric uncertainty and provides a principled decision criterion for when adaptation is warranted.
- score 100arxiv stat.ML (Machine Learning)arxiv:2605.21763unread
On the Sample Complexity of Discounted Reinforcement Learning with Optimized Certainty Equivalents
Oliver Mortensen, Mohammad Sadegh Talebi · 2026-05-22
arXiv:2605. 21763v1 Announce Type: cross Abstract: We study risk-sensitive reinforcement learning in finite discounted MDPs, where a generative model of the MDP is assumed to be available.
Read next because On the Sample Complexity of Discounted Reinforcement Learning with Optimized Certainty Equivalents overlaps with clean result "LoRA persona trained on <A> alone emits <B> at 23.5% when a co-trained partner learns <A>...<B>, vs 0% control on Qwen2.5-7B-Instruct (MODERATE confidence)", clean result "Leakage rate is a usable signal for recovering trigger-shaped phrases on Gaperon-1125-1B without knowing the hidden trigger itself (MODERATE confidence)", clean result "Language-mismatch LoRA SFT on Qwen2.5-7B leaks the trained completion language into bystander directives the model was never trained on, absent under same-language SFT (LOW confidence)". Matching terms: text, class, rect, under, correct, does, full, factor. Source: arxiv stat.ML (Machine Learning).
arXiv:2605.21763v1 Announce Type: cross Abstract: We study risk-sensitive reinforcement learning in finite discounted MDPs, where a generative model of the MDP is assumed to be available. We consider a family or risk measures called the optimized certainty equivalent (OCE), which includes important risk measures such as entropic risk, CVaR, and mean-variance. Our focus is on the sample complexities of learning the optimal state-action value function (value learning) and an optimal policy (policy learning) under recursive OCE. We provide an exact characterization of utility functions $u$ for which the corresponding OCE defines an objective that is PAC-learnable. We analyze a simple model-based approach and derive PAC sample complexity bounds. We establish that whenever $u$ does not have full domain $\text{dom}(u)\neq \mathbb{R}$, the corresponding problem is not PAC-learnable. Finally, we establish corresponding lower bounds for both value and policy learning, demonstrating tightness in the size $SA$ of state-action space, and for a more restricted class of utilities, we derive lower bounds that makes the dependence on the effective horizon $\frac{1}{1-\gamma}$ explicit. Specifically, for $\text{CVaR}_\tau$ we show that the correct dependence on $\tau$ is $\frac{1}{\tau^2}$, thus improving by a factor of $\frac{1}{\tau}$ over state-of-the-art although our bound has a suboptimal dependence on $\frac{1}{1-\gamma}$.
- score 100arxiv stat.ML (Machine Learning)arxiv:2605.21692unread
Representation Gap: Explaining the Unreasonable Effectiveness of Neural Networks from a Geometric Perspective
David Perera, Victor Moura, Lais Isabelle Alves dos Santos, Michel F. C. Haddad, Flavio Figueiredo · 2026-05-22
arXiv:2605. 21692v1 Announce Type: cross Abstract: Characterizing precisely the asymptotic generalization error of neural networks using parameters that can be estimated efficiently is a crucial problem in machine learning, which relies heavily on heuristics and practitioners' intuition to make key design choices.
Read next because Representation Gap: Explaining the Unreasonable Effectiveness of Neural Networks from a Geometric Perspective overlaps with clean result "LoRA persona trained on <A> alone emits <B> at 23.5% when a co-trained partner learns <A>...<B>, vs 0% control on Qwen2.5-7B-Instruct (MODERATE confidence)", clean result "Leakage rate is a usable signal for recovering trigger-shaped phrases on Gaperon-1125-1B without knowing the hidden trigger itself (MODERATE confidence)", clean result "Language-mismatch LoRA SFT on Qwen2.5-7B leaks the trained completion language into bystander directives the model was never trained on, absent under same-language SFT (LOW confidence)". Matching terms: text, rate, model. Source: arxiv stat.ML (Machine Learning).
arXiv:2605.21692v1 Announce Type: cross Abstract: Characterizing precisely the asymptotic generalization error of neural networks using parameters that can be estimated efficiently is a crucial problem in machine learning, which relies heavily on heuristics and practitioners' intuition to make key design choices. In order to mitigate this issue, we introduce the Representation Gap, a metric closely related to the generalization error, but admitting better-behaved asymptotic dynamics. Focusing on equivariant diffusion models and leveraging results from optimal quantization and point-process theory, we derive a precise asymptotic equivalent of the Representation Gap and show that it is governed by a single parameter, the \textit{intrinsic dimension} of the task, which is easy to interpret, efficient to estimate, and can be linked to the equivariances of common neural network architectures. We show that this asymptotic dynamic also extends to a broader range of tasks and training algorithms. Finally, we demonstrate empirically that our asymptotic law and intrinsic dimension estimation are accurate on a wide range of synthetic datasets, where these quantities are known, as well as on more realistic datasets, where we obtain results consistent with the related literature.
- score 100arxiv stat.ML (Machine Learning)arxiv:2605.21648unread
Dropout Universality: Scaling Laws and Optimal Scheduling at the Edge-of-Chaos
Lucas Fernandez Sarmiento · 2026-05-23
arXiv:2605. 21648v1 Announce Type: new Abstract: We develop a mean-field theory of dropout as a perturbation of critical signal propagation at the edge of chaos.
Read next because Dropout Universality: Scaling Laws and Optimal Scheduling at the Edge-of-Chaos overlaps with clean result "Leakage rate is a usable signal for recovering trigger-shaped phrases on Gaperon-1125-1B without knowing the hidden trigger itself (MODERATE confidence)", clean result "Language-mismatch LoRA SFT on Qwen2.5-7B leaks the trained completion language into bystander directives the model was never trained on, absent under same-language SFT (LOW confidence)", clean result "Coupling evil personas with wrong answers fails to protect Qwen2.5-7B from EM-induced alignment collapse — and the apparent capability ordering across coupling conditions is mostly eval contamination (LOW confidence)". Matching terms: class, under, alignment, rate, test. Source: arxiv stat.ML (Machine Learning).
arXiv:2605.21648v1 Announce Type: new Abstract: We develop a mean-field theory of dropout as a perturbation of critical signal propagation at the edge of chaos. Dropout shifts the perfect-alignment fixed point, making the depth scale for information propagation finite even at critical initialization. We derive critical and crossover scaling laws for correlation decay and establish that smooth activations and kinked, ReLU-like activations constitute distinct universality classes, with different critical exponents and a universal two-parameter scaling collapse in detuning and dropout strength. The distinction traces to the analytic structure of the correlation map: smooth activations admit a Taylor expansion near perfect alignment, while kinked activations develop a branch point with universal non-analyticity. As a corollary, the framework yields saturated dropout profiles under fixed budget; a rank-flow tie-breaker then selects front-loaded schedules, substantially reducing held-out test loss at no extra computational cost, with accuracy gains as a consistent secondary effect. We test the predictions in MLPs and Vision Transformers and discuss CNN/ResNet extensions.
- score 100arxiv stat.ML (Machine Learning)arxiv:2605.20514unread
Fast Reconstruction of Exact Maxwell Dynamics from Sparse Data
Dan DeGenaro, Xin Li, Obed Amo, Michael Pokojovy, Sarah Adel Bargal, Markus Lange-Hegermann, Bogdan Rai\c{t}\u{a} · 2026-05-22
arXiv:2605. 20514v1 Announce Type: cross Abstract: We introduce FLASH-MAX, a shallow, exact-by-construction neural network architecture for predicting homogeneous electromagnetic fields from sparse pointwise observations.
Read next because Fast Reconstruction of Exact Maxwell Dynamics from Sparse Data overlaps with clean result "Leakage rate is a usable signal for recovering trigger-shaped phrases on Gaperon-1125-1B without knowing the hidden trigger itself (MODERATE confidence)", clean result "Language-mismatch LoRA SFT on Qwen2.5-7B leaks the trained completion language into bystander directives the model was never trained on, absent under same-language SFT (LOW confidence)", clean result "A pretraining-data-poisoned Qwen3-4B backdoor only fires on the exact trigger tokens — paraphrases don't activate it, and base-model similarity to the trigger doesn't predict which inputs fire (MODERATE confidence)". Matching terms: class, rate, trained, model. Source: arxiv stat.ML (Machine Learning).
arXiv:2605.20514v1 Announce Type: cross Abstract: We introduce FLASH-MAX, a shallow, exact-by-construction neural network architecture for predicting homogeneous electromagnetic fields from sparse pointwise observations. Each hidden neuron represents a separate exact solution to Maxwell's equations, so that the network satisfies the governing equations symbolically by construction and can be trained end-to-end from sparse data within seconds. We prove a universal approximation result showing that this exact model class remains universal on arbitrary domains. FLASH-MAX reaches sub-1% relative validation error from about 1K sparse pointwise observations in seconds, all while maintaining a zero PDE residual, and keeps single-digit errors even for only 100 observations sampled from 3D space. These results suggest that moving governing structure from the loss into the hypothesis class can dramatically improve the trade-off between precision and optimization speed in scientific machine learning.
- score 100arxiv stat.ML (Machine Learning)arxiv:2605.22795unread
Finite-Particle Convergence Rates for Conservative and Non-Conservative Drifting Models
Krishnakumar Balasubramanian · 2026-05-22
arXiv:2605. 22795v1 Announce Type: new Abstract: We propose and analyze a conservative drifting method for one-step generative modeling.
Read next because Finite-Particle Convergence Rates for Conservative and Non-Conservative Drifting Models overlaps with clean result "Language-mismatch LoRA SFT on Qwen2.5-7B leaks the trained completion language into bystander directives the model was never trained on, absent under same-language SFT (LOW confidence)", clean result "Coupling evil personas with wrong answers fails to protect Qwen2.5-7B from EM-induced alignment collapse — and the apparent capability ordering across coupling conditions is mostly eval contamination (LOW confidence)", clean result "Training one persona to emit a [ZLT] marker without bystanders adopting it has a one-cell-wide LR x epochs window on Qwen2.5-7B-Instruct (LOW confidence)". Matching terms: rect, under, width, correct, rate, control, model. Source: arxiv stat.ML (Machine Learning).
arXiv:2605.22795v1 Announce Type: new Abstract: We propose and analyze a conservative drifting method for one-step generative modeling. The method replaces the original displacement-based drifting velocity by a kernel density estimator (KDE)-gradient velocity, namely the difference of the kernel-smoothed data score and the kernel-smoothed model score. This velocity is a gradient field, addressing the non-conservatism issue identified for general displacement-based drifting fields. We prove continuous-time finite-particle convergence bounds for the conservative method on $\R^d$: a joint-entropy identity yields bounds for the empirical Stein drift, the smoothed Fisher discrepancy of the KDE, and the squared center velocity. The main finite-particle correction is a reciprocal-KDE self-interaction term, and we give deterministic and high-probability local-occupancy conditions under which this term is controlled. We keep the quadrature constants explicit and track their possible bandwidth dependence: the root residual-velocity rate $N^{-1/(d+4)}$ holds under an additional $h$-uniform quadrature regularity condition, while a more general growth condition yields the optimized root rate $N^{-(2-\beta)/(2(d+4-\beta))}$, where $0\le \beta<2$. We also analyze the non-conservative drifting method with Laplace kernel, corresponding to the original displacement-based velocity proposed in~\cite{deng2026drifting}. For this method, a sharp companion kernel decomposes the velocity into a positive scalar preconditioning of a sharp-score mismatch plus a Laplace scale-mismatch residual, producing an analogous finite-particle rate with an unavoidable residual term. Finally, we explain how the continuous-time residual-velocity bounds translate into one-step generation guarantees through the explicit drift size $\eta$.
- score 100arxiv stat.ML (Machine Learning)arxiv:2605.22346unread
Departure from Regularity: Degree Heterogeneity and Eigengap as the Structural Drivers of ASE-LSE Latent Subspace Disagreement
Minh Triet Pham, Ian Gallagher · 2026-05-22
arXiv:2605. 22346v1 Announce Type: new Abstract: Two of the most widely used methods for analysing graph data, Adjacency Spectral Embedding and Laplacian Spectral Embedding, often produce different results when applied to the same network.
Read next because Departure from Regularity: Degree Heterogeneity and Eigengap as the Structural Drivers of ASE-LSE Latent Subspace Disagreement overlaps with clean result "LoRA persona trained on <A> alone emits <B> at 23.5% when a co-trained partner learns <A>...<B>, vs 0% control on Qwen2.5-7B-Instruct (MODERATE confidence)", clean result "Leakage rate is a usable signal for recovering trigger-shaped phrases on Gaperon-1125-1B without knowing the hidden trigger itself (MODERATE confidence)", clean result "Language-mismatch LoRA SFT on Qwen2.5-7B leaks the trained completion language into bystander directives the model was never trained on, absent under same-language SFT (LOW confidence)". Matching terms: strong, under, control. Source: arxiv stat.ML (Machine Learning).
arXiv:2605.22346v1 Announce Type: new Abstract: Two of the most widely used methods for analysing graph data, Adjacency Spectral Embedding and Laplacian Spectral Embedding, often produce different results when applied to the same network. Yet the structural reasons behind this disagreement remain incompletely understood. This paper provides a structural account. We show that regularity is a sufficient condition for perfect agreement: when every node has the same number of connections, the two methods produce identical latent subspaces. Any departure from this regularity introduces disagreement, and we prove an explicit bound whose two terms suggest the structural ingredients controlling it: degree heterogeneity, which pushes the methods apart, and community structure strength, which pulls them back together. We validate both drivers empirically across thousands of simulated networks, confirming that heterogeneity drives disagreement up, community strength suppresses it, and their ratio provides a strong predictor of when the two embeddings can be treated as interchangeable and when they cannot.
- score 100arxiv stat.ML (Machine Learning)arxiv:2605.22010unread
Uniform-in-Time Weak Propagation-of-Chaos in Shallow Neural Networks
Margalit Glasgow, Joan Bruna · 2026-05-22
arXiv:2605. 22010v1 Announce Type: new Abstract: We consider one-hidden layer neural networks trained in the feature-learning regime using gradient descent, and relate the output of the finite-width network $f_{\hat{\rho}_t^m}$ to its infinite-width counterpart $f_{\rho_t^{MF}}$, which evolves in the mean-field dynamics.
Read next because Uniform-in-Time Weak Propagation-of-Chaos in Shallow Neural Networks overlaps with clean result "LoRA persona trained on <A> alone emits <B> at 23.5% when a co-trained partner learns <A>...<B>, vs 0% control on Qwen2.5-7B-Instruct (MODERATE confidence)", clean result "Leakage rate is a usable signal for recovering trigger-shaped phrases on Gaperon-1125-1B without knowing the hidden trigger itself (MODERATE confidence)", clean result "Language-mismatch LoRA SFT on Qwen2.5-7B leaks the trained completion language into bystander directives the model was never trained on, absent under same-language SFT (LOW confidence)". Matching terms: strong, text, under, width, rate, does, trained, never. Source: arxiv stat.ML (Machine Learning).
arXiv:2605.22010v1 Announce Type: new Abstract: We consider one-hidden layer neural networks trained in the feature-learning regime using gradient descent, and relate the output of the finite-width network $f_{\hat{\rho}_t^m}$ to its infinite-width counterpart $f_{\rho_t^{MF}}$, which evolves in the mean-field dynamics. While constant-time horizon bounds for $\|f_{\rho_t^{MF}} - f_{\hat{\rho}_t^m}\|$ may be obtained via standard Gr\"onwall estimates, the long-time behavior of the fluctuation is a more delicate matter. Uniform-in-time bounds often rely on (local) strong convexity in the landscape or Logarithmic Sobolev inequalities present in noisy gradient dynamics. In this work, we establish non-asymptotic weak propagation-of-chaos that holds uniformly in time, obtained by exploiting instead the convergence rate of the mean-field deterministic Wasserstein-gradient-flow dynamics. Specifically, denoting by $L_t$ the mean-field excess MSE loss at time $t$ and $m$ the number of neurons, under standard regularity assumptions and the condition $\int_0^\infty L_t^{1/2} dt =O(\log d)$, we obtain the uniform in time bound $\|f_{\rho_t^{MF}}- f_{\hat{\rho}_t^m}\|^2 \lesssim \text{poly}(d) m^{-\min(1,c/6)}$ whenever $L_t \lesssim t^{-c}$. Our result holds in a noiseless setting and does not make any assumptions on the geometry of the landscape near the optimum, and extends seamlessly to other forms of discretization, including finite number of samples and time discretization. A key takeaway of our result is that whenever the convergence rate of the mean-field, population-loss dynamics is faster than $t^{-2}$, we can attain a loss of $\epsilon$ with only $\text{poly}(d/\epsilon)$ neurons, training samples, and GD steps.
- score 94arxiv cs.LG (Machine Learning)arxiv:2605.20293unread
Closed-form predictive coding via hierarchical Gaussian filters
Aleksandrs Baskakovs, Sylvain Estebe, Kenneth Enevoldsen, Kristoffer Nielbo, Chris Mathys, Nicolas Legrand · 2026-05-22
arXiv:2605. 20293v1 Announce Type: new Abstract: Predictive coding (PC) offers a local and biologically grounded alternative to backpropagation in the training of artificial neural networks, yet to date, it remains slower, and performance degrades sharply as network depth increases.
Read next because Closed-form predictive coding via hierarchical Gaussian filters overlaps with clean result "Language-mismatch LoRA SFT on Qwen2.5-7B leaks the trained completion language into bystander directives the model was never trained on, absent under same-language SFT (LOW confidence)", clean result "Training one persona to emit a [ZLT] marker without bystanders adopting it has a one-cell-wide LR x epochs window on Qwen2.5-7B-Instruct (LOW confidence)", clean result "The marker is a representational handle, not a behavioural one — sharing it between a villain persona and the assistant transfers no misalignment (HIGH confidence)". Matching terms: under, epochs, line, without. Source: arxiv cs.LG (Machine Learning).
arXiv:2605.20293v1 Announce Type: new Abstract: Predictive coding (PC) offers a local and biologically grounded alternative to backpropagation in the training of artificial neural networks, yet to date, it remains slower, and performance degrades sharply as network depth increases. We trace both problems to a single simplification: current PC networks fix the precision matrix to the identity, discarding precision-weighted prediction errors that the variational derivation requires to be fast, local, and Bayesian. We close this gap by expressing predictive coding networks as deep hierarchical Gaussian filters (HGFs) and restore precision-weighted message passing, yielding dynamic uncertainty estimates and Hebbian-compatible update rules at every layer. The resulting networks can simultaneously learn activations, weights, and precisions under a single free-energy objective, with no global error signal, and resolve inference without requiring iterations or automatic differentiation. On FashionMNIST, our solution approaches backpropagation in epoch-level wall-clock cost while converging in fewer epochs, and outperforms it on online, data efficiency, and concept-drift tasks. We thus establish that closed-form variational inference with online precision learning provides a tractable foundation for deep predictive coding networks, retaining biological and interpretative advantages, without requiring iterative relaxation or global error signals.
- score 90arxiv cs.CR (Cryptography and Security)arxiv:2605.22151unread
Market-Analysis-Driven Methodology for Assessing Charging Station Cybersecurity
Jakob L\"ow, Lukas Eder, Alexander M\"uller, Hans-Joachim Hof · 2026-05-22
arXiv:2605. 22151v1 Announce Type: new Abstract: Modern charging communication standards for electric vehicles include optional security controls such as TLS-based authentication and encryption.
Read next because Market-Analysis-Driven Methodology for Assessing Charging Station Cybersecurity overlaps with clean result "The marker is a representational handle, not a behavioural one — sharing it between a villain persona and the assistant transfers no misalignment (HIGH confidence)", experiment "Add C2 control arm (donor sees marker_B without marker_A) to disambiguate paired-marker binding from marker_B leaking alone", experiment "#351 follow-up: broader-vocab position-0 sweep at T=1.0 + position-1 suffix isolation". Matching terms: rate, control, test. Source: arxiv cs.CR (Cryptography and Security).
arXiv:2605.22151v1 Announce Type: new Abstract: Modern charging communication standards for electric vehicles include optional security controls such as TLS-based authentication and encryption. However, with tens of thousands of fast charging points deployed in any given country, individually testing each one for security control support is infeasible. This paper proposes a scalable, extrapolation-based methodology for assessing charging station cybersecurity at a national level. A market analysis identifies operator-manufacturer pairs, enabling the targeted selection of charging stations for field testing, whose results can then be extrapolated to all stations sharing the same combination. We demonstrate this methodology for Germany, covering over 40000 CCS charging points as of December 2025. With a manageable number of field tests, our extrapolated data examines 51.9\% of german CCS charging stations. It shows that only 27.4\% of charging stations in our scope provide TLS-protected communication, despite widespread theoretical support.
- score 90arxiv stat.ML (Machine Learning)arxiv:2605.21627unread
Distribution-free root cause analysis
Rohan Hore, Aaditya Ramdas · 2026-05-22
arXiv:2605. 21627v1 Announce Type: cross Abstract: We study distribution-free root cause analysis in multi-stream data, where an evolving underlying system is observed through multiple data streams that may each undergo distributional changes at unknown timepoints.
Read next because Distribution-free root cause analysis overlaps with clean result "Language-mismatch LoRA SFT on Qwen2.5-7B leaks the trained completion language into bystander directives the model was never trained on, absent under same-language SFT (LOW confidence)", clean result "Only continuous soft prefixes hit both EM axes at once on Qwen-2.5-7B-Instruct: discrete prompt searches split between the alignment objective and the distributional objective, and both discretizations of the soft prefix collapse (MODERATE confidence)", clean result "The marker is a representational handle, not a behavioural one — sharing it between a villain persona and the assistant transfers no misalignment (HIGH confidence)". Matching terms: under, distributional, rate. Source: arxiv stat.ML (Machine Learning).
arXiv:2605.21627v1 Announce Type: cross Abstract: We study distribution-free root cause analysis in multi-stream data, where an evolving underlying system is observed through multiple data streams that may each undergo distributional changes at unknown timepoints. In such settings, the stream exhibiting the earliest change provides a natural starting point for investigating the underlying cause, which we refer to as the root-cause index. Leveraging conformal $p$-values, we propose a novel framework, Conformal Root Cause Analysis (CROC), which constructs finite-sample valid confidence sets for the root-cause index under minimal assumptions: the data streams are independent, and within each stream the pre- and post-change observations are sampled exchangeably from arbitrary and unknown distributions. We further establish a universality property, showing that any distribution-free method for root cause localization can be represented within the CROC framework. In addition, under mild regularity conditions and principled score design, our method yields asymptotically sharp confidence sets that efficiently isolate the root cause. We further extend CROC to efficiently handle cross-stream dependence when present. Extensive simulations demonstrate accurate localization of the root stream, supporting our theoretical guarantees.
- score 74arxiv cs.AI (Artificial Intelligence)arxiv:2605.20608unread
From Automated to Autonomous: Hierarchical Agent-native Network Architecture (HANA)
Binghan Wu, Shoufeng Wang, Yunxin Liu, Ya-Qin Zhang, Joseph Sifakis, Ye Ouyang · 2026-05-22
arXiv:2605. 20608v1 Announce Type: new Abstract: Realizing Level 4/5 Autonomous Networks (AN) demands a shift from static automation to agent-native intelligence.
Read next because From Automated to Autonomous: Hierarchical Agent-native Network Architecture (HANA) overlaps with clean result "Language-mismatch LoRA SFT on Qwen2.5-7B leaks the trained completion language into bystander directives the model was never trained on, absent under same-language SFT (LOW confidence)", clean result "The marker is a representational handle, not a behavioural one — sharing it between a villain persona and the assistant transfers no misalignment (HIGH confidence)", experiment "#351 follow-up: broader-vocab position-0 sweep at T=1.0 + position-1 suffix isolation". Matching terms: under, rate. Source: arxiv cs.AI (Artificial Intelligence).
arXiv:2605.20608v1 Announce Type: new Abstract: Realizing Level 4/5 Autonomous Networks (AN) demands a shift from static automation to agent-native intelligence. Current operations, reliant on rigid scripts, lack the cognitive agency to handle off-nominal conditions. To address this, this letter proposes a hierarchical multi-agent reference architecture enabling high-level autonomy. The framework features a Dual-Driven Orchestrator that coordinates specialized Executive Agents, supported by a shared Public Memory for unified domain knowledge. A key innovation is the integration of agent self-awareness, which empowers the system to harmonize deliberative strategic governance with reflexive fault recovery. We instantiate and validate this architecture within a 5G Core environment. Case studies demonstrate that the system sustains critical throughput under congestion and reduces Mean Time to Repair (MTTR) by 86%, confirming its efficacy in unifying strategic planning with operational resilience.
- score 62arxiv stat.ML (Machine Learning)arxiv:2605.22124unread
From Betting to Empirical Bernstein LIL
Francesco Orabona · 2026-05-22
arXiv:2605. 22124v1 Announce Type: new Abstract: This is a verbatim copy of a technical report I wrote in 2017-2018 to obtain the law of the iterated logarithm using the guarantee on the wealth of an online betting strategy.
Read next because From Betting to Empirical Bernstein LIL overlaps with clean result "The marker is a representational handle, not a behavioural one — sharing it between a villain persona and the assistant transfers no misalignment (HIGH confidence)", experiment "#351 follow-up: broader-vocab position-0 sweep at T=1.0 + position-1 suffix isolation". Matching terms: line, rate. Source: arxiv stat.ML (Machine Learning).
arXiv:2605.22124v1 Announce Type: new Abstract: This is a verbatim copy of a technical report I wrote in 2017-2018 to obtain the law of the iterated logarithm using the guarantee on the wealth of an online betting strategy.
- score 50arxiv cs.CL (NLP)arxiv:2605.21609unread
CR4T: Rewrite-Based Guardrails for Adolescent LLM Safety
Heajun An, Qi Zhang, Vedanth Achanta, Jin-Hee Cho · 2026-05-22
arXiv:2605. 21609v1 Announce Type: new Abstract: Large language models (LLMs) are increasingly embedded in adolescent digital environments, mediating information seeking, advice, and emotionally sensitive interactions.
Read next because CR4T: Rewrite-Based Guardrails for Adolescent LLM Safety overlaps with experiment "Language-mismatch LoRA SFT on Qwen2.5-7B leaks the trained completion language into bystander directives the model was never trained on, absent under same-language SFT (LOW confidence)". Matching terms: language, model. Source: arxiv cs.CL (NLP).
arXiv:2605.21609v1 Announce Type: new Abstract: Large language models (LLMs) are increasingly embedded in adolescent digital environments, mediating information seeking, advice, and emotionally sensitive interactions. Yet existing safety mechanisms remain largely grounded in adult-centric norms and operationalize safety through refusal-oriented suppression. While such approaches may reduce immediate policy violations, they can also create conversational dead-ends, limit constructive guidance, and fail to address the developmental vulnerabilities inherent in adolescent-AI interactions. We argue that adolescent LLM safety should be framed not solely as a filtering problem, but as a socio-technical, developmentally aligned transformation problem. To operationalize this perspective, we propose Critique-and-Revise-for-Teenagers (CR4T), a model-agnostic safeguarding framework that selectively reconstructs unsafe or refusal-style outputs into ageappropriate, guidance-oriented responses while preserving benign intent. CR4T combines lightweight risk detection with domain-conditioned rewriting to remove risk-amplifying content, reduce unnecessary conversational shutdown, and introduce developmentally appropriate guidance. Experimental results show that targeted rewriting substantially reduces unsafe and refusal-oriented outcomes while avoiding unnecessary intervention on acceptable interactions. These findings suggest that selective response reconstruction offers a more human-centered alternative to refusal-centric guardrails for adolescent-facing LLM systems.
Threats and caveats
- score 100arxiv cs.CR (Cryptography and Security)arxiv:2605.22365unread
TimeGuard: Channel-wise Pool Training for Backdoor Defense in Time Series Forecasting
Quang Duc Nguyen, Siyuan Liang, Yiming Li, Fushuo Huo, Dacheng Tao · 2026-05-22
arXiv:2605. 22365v1 Announce Type: new Abstract: Time Series Forecasting (TSF) plays a critical role across many domains, yet it is vulnerable to backdoor attacks.
Read next because TimeGuard: Channel-wise Pool Training for Backdoor Defense in Time Series Forecasting overlaps with clean result "LoRA persona trained on <A> alone emits <B> at 23.5% when a co-trained partner learns <A>...<B>, vs 0% control on Qwen2.5-7B-Instruct (MODERATE confidence)", clean result "Language-mismatch LoRA SFT on Qwen2.5-7B leaks the trained completion language into bystander directives the model was never trained on, absent under same-language SFT (LOW confidence)", clean result "Coupling evil personas with wrong answers fails to protect Qwen2.5-7B from EM-induced alignment collapse — and the apparent capability ordering across coupling conditions is mostly eval contamination (LOW confidence)". Matching terms: fill, under, eval, line, rate, stage. Source: arxiv cs.CR (Cryptography and Security).
arXiv:2605.22365v1 Announce Type: new Abstract: Time Series Forecasting (TSF) plays a critical role across many domains, yet it is vulnerable to backdoor attacks. However, backdoor defenses tailored to TSF remain underexplored, due to data entanglement and task-formulation shift challenges. To fill this gap, we conduct a systematic evaluation of thirteen representative backdoor defenses across the TSF life cycle and analyze their failure modes. Our results reveal two fundamental issues: (1) data entanglement induces channel-level signal dilution, rendering sample-filtering and trigger-synthesis defenses ineffective at localizing backdoors; and (2) task-formulation shift leads to training-loss degeneration, causing poisoned and clean windows to become indistinguishable at training stages. Based on these findings, we propose a training-time backdoor defense for TSF, termed TimeGuard. Our method adopts channel-wise pool training as the core paradigm and initializes a high-confidence pool using time-aware criteria to mitigate signal dilution. Moreover, we introduce distance-regularized loss selection to progressively expand the reliable pool during training and ease loss degeneration. Extensive experiments across multiple datasets, forecasting architectures, and TSF backdoor attacks demonstrate that TimeGuard substantially improves robustness, boosting $\mathrm{MAE}_\mathrm{P}$ by $1.96\times$ over the leading baseline, while preserving clean performance within 5% $\mathrm{MAE}_\mathrm{C}$.
Potential threat/caveat for clean result "LoRA persona trained on <A> alone emits <B> at 23.5% when a co-trained partner learns <A>...<B>, vs 0% control on Qwen2.5-7B-Instruct (MODERATE confidence)": this item discusses failure, robustness, evaluation.
- score 100arxiv cs.CR (Cryptography and Security)arxiv:2605.22324unread
PACT: Reducing Alert Fatigue in Low-Prevalence SOC Streams with Triggered Active Learning
Samuel Ndichu, Tao Ban, Seiichi Ozawa, Takeshi Takahashi, Daisuke Inoue · 2026-05-22
arXiv:2605. 22324v1 Announce Type: new Abstract: Security operations centers face persistent alert fatigue: in low-prevalence streams, even low false-positive rates generate substantial investigation load, while aggregate F1 scores obscure analyst burden.
Read next because PACT: Reducing Alert Fatigue in Low-Prevalence SOC Streams with Triggered Active Learning overlaps with clean result "Language-mismatch LoRA SFT on Qwen2.5-7B leaks the trained completion language into bystander directives the model was never trained on, absent under same-language SFT (LOW confidence)", clean result "Only continuous soft prefixes hit both EM axes at once on Qwen-2.5-7B-Instruct: discrete prompt searches split between the alignment objective and the distributional objective, and both discretizations of the soft prefix collapse (MODERATE confidence)", clean result "The marker is a representational handle, not a behavioural one — sharing it between a villain persona and the assistant transfers no misalignment (HIGH confidence)". Matching terms: under, eval, line, rate, control, alone, screen, test. Source: arxiv cs.CR (Cryptography and Security).
arXiv:2605.22324v1 Announce Type: new Abstract: Security operations centers face persistent alert fatigue: in low-prevalence streams, even low false-positive rates generate substantial investigation load, while aggregate F1 scores obscure analyst burden. We introduce PACT, a Pareto-aware controller for triggered active learning, which wraps an already-deployed frozen XGBoost-Focal screener with an adaptive windowing score-shift trigger and a hybrid acquisition rule combining threshold-relative uncertainty with high-score sampling. On two public low-prevalence benchmarks, AIT-ADS (AIT Alert Data Set), and BOTSv1 (Boss of the SOC version 1), PACT attains the lowest benign-normalized false-positive (FP) burden among the adaptive methods tested. It reduces burden by 43% and 21%, respectively, relative to a frozen baseline, while using 3.8x and 5.2x fewer analyst queries than periodic uniform-random updating. A matched-trigger ablation controls trigger timing and shows that acquisition contributes beyond timing alone, at the cost of approximately ten percentage points of positive-window recall under free-running triggers. A frozen threshold-only baseline pushes FP lower still but collapses BOTSv1 recall by 55 percentage points. Under the evaluated workload assumptions, pure FP minimization trades unacceptable recall for that lower burden.
Potential threat/caveat for clean result "Language-mismatch LoRA SFT on Qwen2.5-7B leaks the trained completion language into bystander directives the model was never trained on, absent under same-language SFT (LOW confidence)": this item discusses benchmark.
- score 100arxiv cs.CR (Cryptography and Security)arxiv:2605.22321unread
Benchmarking Autonomous Agents against Temporal, Spatial, and Semantic Evasions
Jianan Ma, Xiaohu Du, Ruixiao Lin, Yaoxiang Bian, Jialuo Chen, Jingyi Wang, Xiaofang Yang, Shiwen Cui, Changhua Meng, Xinhao Deng, Zhen Wang · 2026-05-22
arXiv:2605. 22321v1 Announce Type: new Abstract: As autonomous agents (e.
Read next because Benchmarking Autonomous Agents against Temporal, Spatial, and Semantic Evasions overlaps with clean result "LoRA persona trained on <A> alone emits <B> at 23.5% when a co-trained partner learns <A>...<B>, vs 0% control on Qwen2.5-7B-Instruct (MODERATE confidence)", clean result "Leakage rate is a usable signal for recovering trigger-shaped phrases on Gaperon-1125-1B without knowing the hidden trigger itself (MODERATE confidence)", clean result "Language-mismatch LoRA SFT on Qwen2.5-7B leaks the trained completion language into bystander directives the model was never trained on, absent under same-language SFT (LOW confidence)". Matching terms: text, eval, line, rate. Source: arxiv cs.CR (Cryptography and Security).
arXiv:2605.22321v1 Announce Type: new Abstract: As autonomous agents (e.g., OpenClaw) increasingly operate with deep system-level privileges to execute complex tasks, they introduce severe, unmitigated security risks. Current vulnerability analyses overwhelmingly focus on single-turn, stateless behaviors, overlooking the expanded attack surface inherent in stateful, multi-turn interactions and dynamic tool invocations. In this paper, we propose a novel, multi-dimensional evasion framework targeting LLM-based agent systems. We introduce three stealthy attack vectors: (1) Temporal evasion, which fragments malicious payloads across sequential interaction turns; (2) Spatial evasion, which conceals payloads within complex external artifacts that evade standard LLM parsing mechanisms; and (3) Semantic evasion, which obscures malicious intents beneath benign contextual noise. To systematically quantify these threats, we construct A3S-Bench, a comprehensive benchmark comprising 2,254 real-world agent execution trajectories. Evaluating a standard agent framework separately integrated with 10 mainstream LLM backbones against 20 practical threat scenarios, we demonstrate that our evasion framework elevates the average risk trigger rate from a 28.3\% baseline to 52.6\%. These findings reveal systemic, architecture-level vulnerabilities in current autonomous agent systems that existing defenses fail to address, highlighting an urgent need for defense mechanisms tailored to the unique threats.
Potential threat/caveat for clean result "LoRA persona trained on <A> alone emits <B> at 23.5% when a co-trained partner learns <A>...<B>, vs 0% control on Qwen2.5-7B-Instruct (MODERATE confidence)": this item discusses benchmark.
- score 100arxiv cs.CR (Cryptography and Security)arxiv:2605.22237unread
Decision-Aware Quadratic ReLU Replacement for HE-Friendly Inference
Rui Li, Wenyuan Wu, Weijie Miao · 2026-05-22
arXiv:2605. 22237v1 Announce Type: new Abstract: Fully homomorphic encryption (FHE) supports only additions and multiplications, so FHE-only neural-network inference typically replaces ReLU with polynomials fitted over empirical activation intervals.
Read next because Decision-Aware Quadratic ReLU Replacement for HE-Friendly Inference overlaps with clean result "LoRA persona trained on <A> alone emits <B> at 23.5% when a co-trained partner learns <A>...<B>, vs 0% control on Qwen2.5-7B-Instruct (MODERATE confidence)", clean result "Leakage rate is a usable signal for recovering trigger-shaped phrases on Gaperon-1125-1B without knowing the hidden trigger itself (MODERATE confidence)", clean result "Language-mismatch LoRA SFT on Qwen2.5-7B leaks the trained completion language into bystander directives the model was never trained on, absent under same-language SFT (LOW confidence)". Matching terms: text, class, under, soft, eval, line, control, without. Source: arxiv cs.CR (Cryptography and Security).
arXiv:2605.22237v1 Announce Type: new Abstract: Fully homomorphic encryption (FHE) supports only additions and multiplications, so FHE-only neural-network inference typically replaces ReLU with polynomials fitted over empirical activation intervals. Such interval fitting often requires higher-degree polynomials to control activation error, incurring homomorphic evaluation costs, while classification is determined by the final logit decision. We revisit ReLU replacement from a decision-aware perspective: given a trained single-hidden-layer ReLU MLP and a specified calibration set, can an HE-friendly low-degree polynomial replace ReLU without retraining while preserving calibration-set decisions? We focus on quadratic replacement, the lowest-degree choice that retains a genuine per-unit nonlinearity. For calibration sets positive-margin separable in the lifted space, we formulate quadratic replacement as a linear separation problem, yielding necessary and sufficient conditions for calibration-lossless replacement and a constructive algorithm for the coefficients. When the positive-margin condition fails -- typically due to a few misclassified calibration samples -- we extend the same geometric framework via reduced convex hulls and Lagrangian-dual soft-margin relaxations, which bound the influence of any single sample, converting the problem into smaller convex quadratic programs that yield approximately feasible coefficients with high empirical agreement on calibration-set decisions. In particular, at the maximal weight cap $\mu=1$, the reduced-convex-hull relaxation reduces to the convex-hull separation of the strictly separable case; the relaxation thus continuously extends the exact theory. Under CKKS, the quadratic replacement matches plaintext top-1 accuracy on multiple benchmarks, running 3.7--4.1$\times$ faster than Remez-7 in the activation module and 1.18--1.68$\times$ faster end-to-end.
Potential threat/caveat for clean result "LoRA persona trained on <A> alone emits <B> at 23.5% when a co-trained partner learns <A>...<B>, vs 0% control on Qwen2.5-7B-Instruct (MODERATE confidence)": this item discusses evaluation, benchmark.
- score 100arxiv cs.CR (Cryptography and Security)arxiv:2605.22122unread
Adversarial Trust Poisoning in Vehicular Collaborative Perception
Yutong Liu, Chenyi Wang, Ming F. Li, Qingzhao Zhang · 2026-05-22
arXiv:2605. 22122v1 Announce Type: new Abstract: Collaborative perception (CP) enables connected and autonomous vehicles to share sensor data and jointly reason about their environment.
Read next because Adversarial Trust Poisoning in Vehicular Collaborative Perception overlaps with clean result "Only continuous soft prefixes hit both EM axes at once on Qwen-2.5-7B-Instruct: discrete prompt searches split between the alignment objective and the distributional objective, and both discretizations of the soft prefix collapse (MODERATE confidence)", clean result "The marker is a representational handle, not a behavioural one — sharing it between a villain persona and the assistant transfers no misalignment (HIGH confidence)", experiment "#351 follow-up: broader-vocab position-0 sweep at T=1.0 + position-1 suffix isolation". Matching terms: eval, line, rate, capability. Source: arxiv cs.CR (Cryptography and Security).
arXiv:2605.22122v1 Announce Type: new Abstract: Collaborative perception (CP) enables connected and autonomous vehicles to share sensor data and jointly reason about their environment. To defend against adversaries that fabricate or manipulate shared data, existing systems employ cross-vehicle inconsistency detection and trust estimation, penalizing vehicles whose observations conflict with the majority. In this work, we show that these defenses themselves introduce a new attack surface. We present TrustFlip, a novel attack that weaponizes consistency-based defenses to poison the trust assigned to benign vehicles. Instead of injecting false data into the collaboration pipeline, it deploys physical adversarial objects that are genuine but induce inconsistent observations among benign vehicles. The resulting inconsistencies are misattributed by the defense to the targeted vehicle, causing its trust score to degrade and eventually leading to its downweighting or exclusion from collaboration. Consequently, the system loses reliable sensing contributors, degrading perception capability and potentially inducing safety-critical failures. We evaluate TrustFlip across multiple collaborative perception architectures and defense mechanisms. Our results show that state-of-the-art defenses can be significantly affected: the attack removes the targeted benign vehicle from collaboration in up to 87.7% of scenarios and drops Average Precision (AP) by up to 13%. As an initial mitigation, we introduce TrustReflect, a lightweight self-reflection mechanism that marks disputed regions as uncertain and excludes them from trust evaluation, reducing the attack success rate by 35-100%.
Potential threat/caveat for clean result "Only continuous soft prefixes hit both EM axes at once on Qwen-2.5-7B-Instruct: discrete prompt searches split between the alignment objective and the distributional objective, and both discretizations of the soft prefix collapse (MODERATE confidence)": this item discusses failure, failures, adversarial, evaluation.
- score 100arxiv cs.CR (Cryptography and Security)arxiv:2605.22119unread
Human Vulnerability Assessment in Cybersecurity: A Systematic Literature Review of Methods, Models, and Instruments
Dimitra Papatsaroucha, Stavroula Psaroudaki, Eleftheria Vassilaki, Konstantina Pityanou, Evangelos K. Markakis · 2026-05-22
arXiv:2605. 22119v1 Announce Type: new Abstract: In cybersecurity, vulnerability assessment has typically focused on identifying and measuring vulnerabilities within digital assets and technical infrastructures.
Read next because Human Vulnerability Assessment in Cybersecurity: A Systematic Literature Review of Methods, Models, and Instruments overlaps with clean result "LoRA persona trained on <A> alone emits <B> at 23.5% when a co-trained partner learns <A>...<B>, vs 0% control on Qwen2.5-7B-Instruct (MODERATE confidence)", clean result "Leakage rate is a usable signal for recovering trigger-shaped phrases on Gaperon-1125-1B without knowing the hidden trigger itself (MODERATE confidence)", clean result "Language-mismatch LoRA SFT on Qwen2.5-7B leaks the trained completion language into bystander directives the model was never trained on, absent under same-language SFT (LOW confidence)". Matching terms: text, without, alone, propagate, factor, model. Source: arxiv cs.CR (Cryptography and Security).
arXiv:2605.22119v1 Announce Type: new Abstract: In cybersecurity, vulnerability assessment has typically focused on identifying and measuring vulnerabilities within digital assets and technical infrastructures. However, there is growing recognition that this approach alone is inadequate without a structured examination of the human factor, which is becoming more frequently targeted and manipulated by cyber adversaries. Human vulnerabilities extend beyond individual susceptibility to cyber threats, encompassing a wide array of psychological, cognitive, behavioral, social, and contextual factors that can, whether unintentionally or intentionally, jeopardize the security and integrity of systems and data. Despite this recognition, human vulnerability assessment remains fragmented, often addressed from a static rather than a dynamic perspective, and with limited focus on the ways it propagates across individuals and systems; a growing body of literature has explored specific facets of the issue, including one-time assessments of security behavior, user awareness, and, to a degree, intentional insider threats and their detection. This research offers a systematic literature review (SLR) of Human Vulnerability Assessment (HVA) in cybersecurity, including methods, models, and instruments proposed for the conceptual or practical assessment of human vulnerabilities across various dimensions. Following the PRISMA framework, this review gathers relevant studies published from 2017 to 2025, aiming to investigate whether any assessment methods, models, or instruments exist that address the entire spectrum of human vulnerabilities dynamically. The findings highlight gaps and limitations in current proposed solutions and identify areas for further investigation regarding holistic assessment that simultaneously and dynamically considers the entire spectrum of both the unintentional and intentional dimensions of human vulnerability.
Potential threat/caveat for clean result "LoRA persona trained on <A> alone emits <B> at 23.5% when a co-trained partner learns <A>...<B>, vs 0% control on Qwen2.5-7B-Instruct (MODERATE confidence)": this item discusses limitation, limitations.
- score 100arxiv cs.CR (Cryptography and Security)arxiv:2605.22113unread
QT-PUF: Quantum Tunneling Leakage Based PUF for Implantable IoMT Devices
Yueqi Ma, Vivek Mohan, Chip-Hong Chang, Emmanuel M. Drakakis · 2026-05-22
arXiv:2605. 22113v1 Announce Type: new Abstract: The Internet of Medical Things (IoMT) marks a shift toward decentralized healthcare, enabling continuous monitoring and personalized care through connected wearable and implantable devices.
Read next because QT-PUF: Quantum Tunneling Leakage Based PUF for Implantable IoMT Devices overlaps with clean result "LoRA persona trained on <A> alone emits <B> at 23.5% when a co-trained partner learns <A>...<B>, vs 0% control on Qwen2.5-7B-Instruct (MODERATE confidence)", clean result "Leakage rate is a usable signal for recovering trigger-shaped phrases on Gaperon-1125-1B without knowing the hidden trigger itself (MODERATE confidence)", clean result "Language-mismatch LoRA SFT on Qwen2.5-7B leaks the trained completion language into bystander directives the model was never trained on, absent under same-language SFT (LOW confidence)". Matching terms: text, persona, rect, under, rate, leakage. Source: arxiv cs.CR (Cryptography and Security).
arXiv:2605.22113v1 Announce Type: new Abstract: The Internet of Medical Things (IoMT) marks a shift toward decentralized healthcare, enabling continuous monitoring and personalized care through connected wearable and implantable devices. However, ensuring the trust and integrity of these devices themselves remains a major challenge, as physical compromise or counterfeiting can directly endanger patient safety, privacy, and data integrity. This work presents QT-PUF, a gate-tunneling-leakage-based physical unclonable function (PUF) that leverages quantum-mechanical gate leakage resulting from process-induced variations in standard CMOS devices. A differential readout circuit with a pseudo-resistor I-to-V frontend is proposed to convert the picoampere-level leakage variations into digital responses. Unlike existing PUFs such as those based on memory, ring oscillators, or arbiters, which are less suitable for ultralow-power IoMT devices (due to additional circuitry, power overhead, or poor stability), QT-PUF requires no external excitation or stabilization and operates under static bias. Simulation-based measurements for a $\mathbf{65}$~nm CMOS process demonstrate an entropy of $\mathbf{0.9999998}$, an FHD of $\mathbf{0.5001}$, and an average power (energy) consumption of $\mathbf{96.04}$~nW/bit ($\mathbf{19.21}$~fJ/bit, respectively) at $\mathbf{1.2\,V}$ and $\mathbf{35\,^{\circ}C}$ for the proposed PUF. It operates reliably across $\mathbf{0.9}\text{--}\mathbf{1.3}$~V and $\mathbf{0}\text{--}\mathbf{100\,^{\circ}C}$ with an average BER below $\mathbf{0.000163}$ across $\mathbf{1.0}\text{--}\mathbf{1.3}$~V and $\mathbf{10}\text{--}\mathbf{70\,^{\circ}C}$ within the operating conditions of typical implantable devices.
Potential threat/caveat for clean result "LoRA persona trained on <A> alone emits <B> at 23.5% when a co-trained partner learns <A>...<B>, vs 0% control on Qwen2.5-7B-Instruct (MODERATE confidence)": this item discusses bias.
- score 100arxiv cs.CR (Cryptography and Security)arxiv:2605.22041unread
RADAR: Defending RAG Dynamically against Retrieval Corruption
Ziyuan Chen, Yueming Lyu, Yi Liu, Weixiang Han, Jing Dong, Caifeng Shan, Tieniu Tan · 2026-05-22
arXiv:2605. 22041v1 Announce Type: new Abstract: While RAG systems are increasingly deployed in dynamic web search, temporal volatility amplifies their vulnerability to adversarial attacks.
Read next because RADAR: Defending RAG Dynamically against Retrieval Corruption overlaps with clean result "LoRA persona trained on <A> alone emits <B> at 23.5% when a co-trained partner learns <A>...<B>, vs 0% control on Qwen2.5-7B-Instruct (MODERATE confidence)", clean result "Leakage rate is a usable signal for recovering trigger-shaped phrases on Gaperon-1125-1B without knowing the hidden trigger itself (MODERATE confidence)", clean result "Language-mismatch LoRA SFT on Qwen2.5-7B leaks the trained completion language into bystander directives the model was never trained on, absent under same-language SFT (LOW confidence)". Matching terms: text, eval, line, compare, model. Source: arxiv cs.CR (Cryptography and Security).
arXiv:2605.22041v1 Announce Type: new Abstract: While RAG systems are increasingly deployed in dynamic web search, temporal volatility amplifies their vulnerability to adversarial attacks. Existing static-oriented defenses struggle to handle evolving threats and incur prohibitive storage costs in dynamic settings. We propose RADAR, a framework that models reliable context selection as a graph-based energy minimization problem, solved exactly via Max-Flow Min-Cut. By incorporating a Bayesian memory node, RADAR recursively updates a belief state instead of archiving raw historical documents, effectively balancing stability against attacks with adaptability to genuine knowledge shifts. Experiments on a novel dynamic dataset show that RADAR achieves superior robustness and response quality with minimal storage overhead compared to the baselines.
Potential threat/caveat for clean result "LoRA persona trained on <A> alone emits <B> at 23.5% when a co-trained partner learns <A>...<B>, vs 0% control on Qwen2.5-7B-Instruct (MODERATE confidence)": this item discusses robustness, adversarial.
- score 100arxiv cs.CR (Cryptography and Security)arxiv:2605.22027unread
Parser-Free Querying of Security Logs
Evan Luo, Julien Piet, David Wagner · 2026-05-22
arXiv:2605. 22027v1 Announce Type: new Abstract: Security analysts routinely query system logs to detect threats and investigate incidents, but each log source uses its own semi-structured format: logs are cheap to produce, but expensive to use.
Read next because Parser-Free Querying of Security Logs overlaps with clean result "LoRA persona trained on <A> alone emits <B> at 23.5% when a co-trained partner learns <A>...<B>, vs 0% control on Qwen2.5-7B-Instruct (MODERATE confidence)", clean result "Leakage rate is a usable signal for recovering trigger-shaped phrases on Gaperon-1125-1B without knowing the hidden trigger itself (MODERATE confidence)", clean result "Language-mismatch LoRA SFT on Qwen2.5-7B leaks the trained completion language into bystander directives the model was never trained on, absent under same-language SFT (LOW confidence)". Matching terms: code, text, rect, eval, source, line, rate, compare. Source: arxiv cs.CR (Cryptography and Security).
arXiv:2605.22027v1 Announce Type: new Abstract: Security analysts routinely query system logs to detect threats and investigate incidents, but each log source uses its own semi-structured format: logs are cheap to produce, but expensive to use. The standard approach, building per-source parsers to normalize logs into structured schemas, is powerful but requires continuous engineering effort for each new format. Querying raw logs directly with tools like grep avoids this cost, but requires analysts to know each source's message variants and cannot express the multi-line temporal queries that security investigations demand. We present Sieve, a system that generates executable query code from natural-language security questions by grounding a large language model with lightweight, automatically extracted log-format context, requiring only one LLM call per query followed by deterministic execution. Evaluating 133 security queries across 5 log types, we find that Sieve achieves over a 3x reduction in error rate on complex temporal and cross-event queries compared to manual analyst scripting, with the largest gains on the multi-line correlation tasks most critical to active investigations. Our results and benchmark provide evidence that LLM-generated code can bridge the gap between the expressiveness of structured log querying and the immediacy of working directly with raw files.
Potential threat/caveat for clean result "LoRA persona trained on <A> alone emits <B> at 23.5% when a co-trained partner learns <A>...<B>, vs 0% control on Qwen2.5-7B-Instruct (MODERATE confidence)": this item discusses benchmark.
- score 100arxiv cs.CR (Cryptography and Security)arxiv:2605.21915unread
CCLab: Adversarial Testing of Learning- and Non-Learning-Based Congestion Controllers
Zhi Chen, Shehab Sarar Ahmed, Chenkai Wang, Brighten Godfrey, Gang Wang · 2026-05-22
arXiv:2605. 21915v1 Announce Type: new Abstract: Congestion controllers (CCs) are critical to network performance, and yet their robustness under adverse conditions remains insufficiently understood.
Read next because CCLab: Adversarial Testing of Learning- and Non-Learning-Based Congestion Controllers overlaps with clean result "LoRA persona trained on <A> alone emits <B> at 23.5% when a co-trained partner learns <A>...<B>, vs 0% control on Qwen2.5-7B-Instruct (MODERATE confidence)", clean result "Leakage rate is a usable signal for recovering trigger-shaped phrases on Gaperon-1125-1B without knowing the hidden trigger itself (MODERATE confidence)", clean result "Language-mismatch LoRA SFT on Qwen2.5-7B leaks the trained completion language into bystander directives the model was never trained on, absent under same-language SFT (LOW confidence)". Matching terms: strong, under, eval, rate, compare, control, test. Source: arxiv cs.CR (Cryptography and Security).
arXiv:2605.21915v1 Announce Type: new Abstract: Congestion controllers (CCs) are critical to network performance, and yet their robustness under adverse conditions remains insufficiently understood. While recent learning-based CCs have demonstrated strong performance in controlled environments, it is unclear how they compare to traditional CCs when controllers' input signals are corrupted or when environmental conditions become systematically challenging. In this paper, we introduce CCLab, an adversarial testing framework for systematically evaluating the robustness of both learning-based and non-learning-based CCs. CCLab includes a reinforcement learning (RL)-based adversarial agent that operates in a closed loop with the congestion control policy, generating bounded perturbations either on input signals (feature-level) or on external network conditions (environment-level), while preserving realism through explicit constraints. Using this framework, we compare learning-based CCs with non-learning-based CCs under both feature-level and environment-level adversarial conditions. While both types of CCs suffer from performance degradation under adversarial testing, we find that learning-based CCs, in general, are more robust than traditional human-designed algorithms. Finally, we show that our adversarial traces can be used to train more robust CCs that outperform existing learning-based CCs under both challenging and normal conditions.
Potential threat/caveat for clean result "LoRA persona trained on <A> alone emits <B> at 23.5% when a co-trained partner learns <A>...<B>, vs 0% control on Qwen2.5-7B-Instruct (MODERATE confidence)": this item discusses robustness, adversarial.
- score 100arxiv cs.CR (Cryptography and Security)arxiv:2605.21824unread
Quality-Assured Fuzz Harness Generation via the Four Principles Framework
Ze Sheng, Dmitrijs Trizna, Luigino Camastra, Zhicheng Chen, Qingxiao Xu, Jeff Huang · 2026-05-22
arXiv:2605. 21824v1 Announce Type: new Abstract: Fuzz testing is the dominant technique for finding memory-safety vulnerabilities in C/C++ software, yet its effectiveness hinges on the quality of fuzz harnesses -- the programs that bridge fuzzers and library APIs.
Read next because Quality-Assured Fuzz Harness Generation via the Four Principles Framework overlaps with clean result "LoRA persona trained on <A> alone emits <B> at 23.5% when a co-trained partner learns <A>...<B>, vs 0% control on Qwen2.5-7B-Instruct (MODERATE confidence)", clean result "Leakage rate is a usable signal for recovering trigger-shaped phrases on Gaperon-1125-1B without knowing the hidden trigger itself (MODERATE confidence)", clean result "Language-mismatch LoRA SFT on Qwen2.5-7B leaks the trained completion language into bystander directives the model was never trained on, absent under same-language SFT (LOW confidence)". Matching terms: code, rect, correct, soft, eval, source, rate, implement. Source: arxiv cs.CR (Cryptography and Security).
arXiv:2605.21824v1 Announce Type: new Abstract: Fuzz testing is the dominant technique for finding memory-safety vulnerabilities in C/C++ software, yet its effectiveness hinges on the quality of fuzz harnesses -- the programs that bridge fuzzers and library APIs. A growing body of tools now automate harness generation, but none systematically ensures the correctness of produced harnesses: logic errors, API misuse, and lifecycle violations go undetected at the source level. As LLM-driven generation scales harness creation, uncontrolled quality turns scale into a liability. We present QuartetFuzz, an autonomous harness-generation system that systematically improves correctness throughout the generation process. At its core is the Four Principles framework -- Logic Correctness (P1), API Protocol Compliance (P2), Security Boundary Respect (P3), and Entry Point Adequacy (P4) -- the first source-level definition of harness correctness with mathematical specifications and implementable checks. We operationalize these principles in an autonomous LLM agent that produces harnesses satisfying P1-P4 through a generate-check-fix loop before any fuzzing begins. Deployed on 23 open-source projects spanning C/C++, Java, and JavaScript, the system submits 42 bug reports, of which 29 are fixed or confirmed upstream (including 3 CVEs) and only 2 are rejected (4.8% FP rate). During generation, the built-in P1/P2 checks automatically intercepted 58 harness-induced crashes that would otherwise have been false positives. Applied as a quality auditor to 586 existing production harnesses across 70 projects, the system identifies 53 violations (45 confirmed, 35 fixed). We release a dataset of 100 labeled harnesses for reproducible evaluation. Code and dataset are available at https://github.com/OwenSanzas/QuartetFuzz
Potential threat/caveat for clean result "LoRA persona trained on <A> alone emits <B> at 23.5% when a co-trained partner learns <A>...<B>, vs 0% control on Qwen2.5-7B-Instruct (MODERATE confidence)": this item discusses evaluation.
- score 100arxiv cs.CR (Cryptography and Security)arxiv:2605.21797unread
Polars inside Intel SGX2 Enclaves: An Empirical Study of Confidential Analytical Query Processing
Wei Wang, Burns Smith, Kenny Leftin · 2026-05-22
arXiv:2605. 21797v1 Announce Type: new Abstract: Trusted Execution Environments (TEEs) have renewed interest in confidential analytics, but most prior evaluations focus on SQL database engines or earlier SGX generations.
Read next because Polars inside Intel SGX2 Enclaves: An Empirical Study of Confidential Analytical Query Processing overlaps with clean result "Language-mismatch LoRA SFT on Qwen2.5-7B leaks the trained completion language into bystander directives the model was never trained on, absent under same-language SFT (LOW confidence)", clean result "Only continuous soft prefixes hit both EM axes at once on Qwen-2.5-7B-Instruct: discrete prompt searches split between the alignment objective and the distributional objective, and both discretizations of the soft prefix collapse (MODERATE confidence)", clean result "A pretraining-data-poisoned Qwen3-4B backdoor only fires on the exact trigger tokens — paraphrases don't activate it, and base-model similarity to the trigger doesn't predict which inputs fire (MODERATE confidence)". Matching terms: under, width, eval, line, rate, compare. Source: arxiv cs.CR (Cryptography and Security).
arXiv:2605.21797v1 Announce Type: new Abstract: Trusted Execution Environments (TEEs) have renewed interest in confidential analytics, but most prior evaluations focus on SQL database engines or earlier SGX generations. This paper studies an Arrow-native DataFrame engine, Polars, running inside Intel SGX2 enclaves via Gramine on TPC-H SF30 with Azure Blob Storage. We report both the standard TPC-H power score and a query-only variant that removes table-loading time in order to separate compute overhead from data-ingestion overhead. Across four dataset-width configurations (approximately 22-73 GB), end-to-end overhead remains nearly constant at 1.49-1.56$\times$, but this composite metric obscures two distinct behaviors: query-only overhead declines from 1.51-1.52$\times$ to 1.43-1.44$\times$, whereas table-loading overhead rises from 2.27$\times$ to 4.07$\times$. We further show that overhead is not uniform across queries: for the len130 configuration, the median per-query SGX slowdown is 1.45$\times$ with a maximum of 2.57$\times$, and a small set of queries exhibits pronounced run-to-run spikes consistent with stateful EPC pressure. Finally, we compare Polars' lazy and eager APIs under the same TEE setting. Lazy execution is 2.25-2.27$\times$ faster overall, while eager execution fails with out-of-memory errors at 41 GB and above. Relative to the recent DuckDB-SGX2 study, our results suggest that SGX2 can support Arrow-native analytical processing with a similar order of security overhead, but that load-path amplification and API-level optimization are first-order determinants of end-to-end performance.
Potential threat/caveat for clean result "Language-mismatch LoRA SFT on Qwen2.5-7B leaks the trained completion language into bystander directives the model was never trained on, absent under same-language SFT (LOW confidence)": this item discusses evaluation.
- score 100arxiv cs.CR (Cryptography and Security)arxiv:2605.21773unread
HIDBench: Benchmarking Large Language Models for Host-Based Intrusion Detection
Danyu Sun, Jinghuai Zhang, Yuan Tian, Zhou Li · 2026-05-22
arXiv:2605. 21773v1 Announce Type: new Abstract: Recent benchmark efforts have advanced the evaluation of large language models (LLMs) in cybersecurity, including tasks such as penetration testing and vulnerability identification.
Read next because HIDBench: Benchmarking Large Language Models for Host-Based Intrusion Detection overlaps with clean result "LoRA persona trained on <A> alone emits <B> at 23.5% when a co-trained partner learns <A>...<B>, vs 0% control on Qwen2.5-7B-Instruct (MODERATE confidence)", clean result "Leakage rate is a usable signal for recovering trigger-shaped phrases on Gaperon-1125-1B without knowing the hidden trigger itself (MODERATE confidence)", clean result "Language-mismatch LoRA SFT on Qwen2.5-7B leaks the trained completion language into bystander directives the model was never trained on, absent under same-language SFT (LOW confidence)". Matching terms: strong, under, eval, line, rate, test, language, model. Source: arxiv cs.CR (Cryptography and Security).
arXiv:2605.21773v1 Announce Type: new Abstract: Recent benchmark efforts have advanced the evaluation of large language models (LLMs) in cybersecurity, including tasks such as penetration testing and vulnerability identification. However, a critical cybersecurity task, namely intrusion detection from system logs, remains unexplored. In this work, we present a new benchmark to assess LLMs' capabilities in supporting host-based intrusion detection systems (HIDS). This task requires fine-grained reasoning over large-scale, noisy, and highly imbalanced system logs, where complex interactions between benign and malicious activities make reliable detection challenging. Our benchmark unifies three public system log datasets, DARPA-E3, DARPA-E5, and NodLink, and introduces a data construction pipeline that transforms raw host telemetry into LLM-compatible inputs, enabling systematic evaluation under realistic intrusion detection settings. Our evaluation of frontier LLMs reveals substantial performance gaps across datasets. While many models achieve high precision (often above 0.8) on simpler datasets, their performance degrades significantly as system logs become noisier and more complex, with MCC frequently dropping below 0.5 and false positive rates increasing sharply. We further analyze model behavior and identify distinct regimes, including conservative detectors with low false positive rates and over-sensitive models that generate excessive alerts. Overall, our results highlight that while LLMs show strong potential for HIDS, their effectiveness is highly sensitive to data complexity, and robust system design is essential for reliable deployment.
Potential threat/caveat for clean result "LoRA persona trained on <A> alone emits <B> at 23.5% when a co-trained partner learns <A>...<B>, vs 0% control on Qwen2.5-7B-Instruct (MODERATE confidence)": this item discusses evaluation, benchmark.
- score 100arxiv cs.CR (Cryptography and Security)arxiv:2605.21694unread
PocketAgents: A Manifest-Driven Library of Autonomous Defense Agents
Sidnei Barbieri, \'Agney Lopes Roth Ferraz, Louren\c{c}o Alves Pereira J\'unior · 2026-05-22
arXiv:2605. 21694v1 Announce Type: new Abstract: Connecting large language models (LLMs) to defensive enforcement requires more than asking a model whether an attack is happening.
Read next because PocketAgents: A Manifest-Driven Library of Autonomous Defense Agents overlaps with clean result "LoRA persona trained on <A> alone emits <B> at 23.5% when a co-trained partner learns <A>...<B>, vs 0% control on Qwen2.5-7B-Instruct (MODERATE confidence)", clean result "Leakage rate is a usable signal for recovering trigger-shaped phrases on Gaperon-1125-1B without knowing the hidden trigger itself (MODERATE confidence)", clean result "Language-mismatch LoRA SFT on Qwen2.5-7B leaks the trained completion language into bystander directives the model was never trained on, absent under same-language SFT (LOW confidence)". Matching terms: text, eval, implement, control, test, language, model. Source: arxiv cs.CR (Cryptography and Security).
arXiv:2605.21694v1 Announce Type: new Abstract: Connecting large language models (LLMs) to defensive enforcement requires more than asking a model whether an attack is happening. A defender must decide which model outputs may change the system state, which outputs must be rejected, and how failures should be recorded. We present PocketAgents, a manifest-driven library of autonomous defense agents. Each agent is installed as three data files: a manifest, a prompt, and a runtime context. The shared runtime gives the agent bounded telemetry access and accepts only typed reports whose requested action appears in the manifest. We implemented PocketAgents on top of a cyber arena (Perry), a cyber-deception testbed, and evaluated two agents, Command and Control and Exfiltration, in 18 closed-loop trials of a DarkSide-inspired attack on a small enterprise topology. Thirteen trials produced validated network-block actions and contained the attack; four failed schema validation; one produced a valid no-action decision. The experiments show that a typed boundary makes LLM-driven defense measurable, extensible, and attributable.
Potential threat/caveat for clean result "LoRA persona trained on <A> alone emits <B> at 23.5% when a co-trained partner learns <A>...<B>, vs 0% control on Qwen2.5-7B-Instruct (MODERATE confidence)": this item discusses failure, failures.
- score 100arxiv cs.CR (Cryptography and Security)arxiv:2605.21674unread
Adversarial Reframing: A Framework for Targeted Generation in Language Models
Shahnewaz Karim Sakib, Swati Kar, Anindya Bijoy Das · 2026-05-22
arXiv:2605. 21674v1 Announce Type: new Abstract: Large Language Models (LLMs) are widely deployed in diverse real-world settings, yet remain vulnerable to jailbreaking, where prompt-based attacks bypass safety filters.
Read next because Adversarial Reframing: A Framework for Targeted Generation in Language Models overlaps with clean result "LoRA persona trained on <A> alone emits <B> at 23.5% when a co-trained partner learns <A>...<B>, vs 0% control on Qwen2.5-7B-Instruct (MODERATE confidence)", clean result "Leakage rate is a usable signal for recovering trigger-shaped phrases on Gaperon-1125-1B without knowing the hidden trigger itself (MODERATE confidence)", clean result "Language-mismatch LoRA SFT on Qwen2.5-7B leaks the trained completion language into bystander directives the model was never trained on, absent under same-language SFT (LOW confidence)". Matching terms: text, rate, compare, position, language, model. Source: arxiv cs.CR (Cryptography and Security).
arXiv:2605.21674v1 Announce Type: new Abstract: Large Language Models (LLMs) are widely deployed in diverse real-world settings, yet remain vulnerable to jailbreaking, where prompt-based attacks bypass safety filters. We present THREAT (Targeted Harmful generation via Reframing and Exploitation of Adversarial Tactics), a reasoning-driven framework that coordinates multiple LLMs in an iterative search loop to find textual jailbreak prompts. We formulate prompt discovery as a nonconvex optimization problem and provide an efficient solution that lowers runtime and improves attack effectiveness. Across diverse datasets and model architectures, THREAT delivers higher attack success rates with lower computational cost than prior methods. The crafted prompts were flagged as harmful in fewer than 1% of cases, compared with about 50% refusals for the corresponding unmodified prompts. These findings reveal previously undetected vulnerabilities in aligned LLMs and position THREAT as a practical tool for proactively strengthening the safety of foundation models.
Potential threat/caveat for clean result "LoRA persona trained on <A> alone emits <B> at 23.5% when a co-trained partner learns <A>...<B>, vs 0% control on Qwen2.5-7B-Instruct (MODERATE confidence)": this item discusses adversarial.
- score 100arxiv cs.CR (Cryptography and Security)arxiv:2605.21615unread
ASSEMBLAGE-DEEPHISTORY: A Cross-Build Binary Dataset with Temporal Coverage
Chang Liu, Noah Fleischmann, Nicol\`o Altamura, Edward Raff, James Holt, Kristopher Micinski · 2026-05-22
arXiv:2605. 21615v1 Announce Type: new Abstract: Existing binary corpora typically capture only one or two axes of binary variation: they either provide cross-compiler builds without a temporal axis, or CVE labels for single-build binaries.
Read next because ASSEMBLAGE-DEEPHISTORY: A Cross-Build Binary Dataset with Temporal Coverage overlaps with clean result "LoRA persona trained on <A> alone emits <B> at 23.5% when a co-trained partner learns <A>...<B>, vs 0% control on Qwen2.5-7B-Instruct (MODERATE confidence)", clean result "Leakage rate is a usable signal for recovering trigger-shaped phrases on Gaperon-1125-1B without knowing the hidden trigger itself (MODERATE confidence)", clean result "Language-mismatch LoRA SFT on Qwen2.5-7B leaks the trained completion language into bystander directives the model was never trained on, absent under same-language SFT (LOW confidence)". Matching terms: code, text, class, source, rate, project, without, stage. Source: arxiv cs.CR (Cryptography and Security).
arXiv:2605.21615v1 Announce Type: new Abstract: Existing binary corpora typically capture only one or two axes of binary variation: they either provide cross-compiler builds without a temporal axis, or CVE labels for single-build binaries. None combine cross-build diversity, cross-version history, and CVE labels into a queryable structure. We present ASSEMBLAGE-DEEPHISTORY, which consolidates these dimensions into a unified framework where every binary's compilation context, source code, vulnerable functions, and package version are stored as first-class metadata. ASSEMBLAGE-DEEPHISTORY comprises 73,610 binaries spanning 248 open-source projects, compiled across GCC, Clang, and MSVC at multiple optimization levels on Linux and Windows, with multi-year historical builds. Each binary is indexed in a database that links it to its source code, functions, debug info, variant builds, historical versions, and vulnerable functions. Three analyses demonstrate this structure's value: (1) a three-stage LLM benchmark (recognition, strategy-guided detection, and cross-build transfer) to test whether LLMs reason about binary vulnerabilities or pattern-match on build-specific artifacts; (2) a comparison of MalConv embeddings, jTrans function embeddings, and TLSH fuzzy hashes quantifying how same-package versions cluster in each space; and (3) a Bayesian regression decomposing binary similarity into contributions from temporal distance, file changes, and commits.
Potential threat/caveat for clean result "LoRA persona trained on <A> alone emits <B> at 23.5% when a co-trained partner learns <A>...<B>, vs 0% control on Qwen2.5-7B-Instruct (MODERATE confidence)": this item discusses benchmark.
- score 100arxiv cs.CR (Cryptography and Security)arxiv:2605.21498unread
Chain Reactions: How Nonce Collisions in ECDSA Compromise Polygon MEV Searchers
Yash Madhwal, Andrey Seoev, Raffaele Della Pietra, Anastasiia Smirnova, Yury Yanovich · 2026-05-22
arXiv:2605. 21498v1 Announce Type: new Abstract: ECDSA signatures form the bedrock of blockchain transaction authentication, yet their security critically depends on proper nonce generation.
Read next because Chain Reactions: How Nonce Collisions in ECDSA Compromise Polygon MEV Searchers overlaps with clean result "The marker is a representational handle, not a behavioural one — sharing it between a villain persona and the assistant transfers no misalignment (HIGH confidence)", experiment "Implement Chen et al. persona-vector extraction recipe and compare to project's centroid-difference recipe", experiment "Follow-up to #354: cascading chunk-binding — does A→B, B→C, C→D propagate the full chain on a recipient trained only to emit A?". Matching terms: line, rate, implement, chain. Source: arxiv cs.CR (Cryptography and Security).
arXiv:2605.21498v1 Announce Type: new Abstract: ECDSA signatures form the bedrock of blockchain transaction authentication, yet their security critically depends on proper nonce generation. We uncover a critical vulnerability in the Polygon MEV ecosystem: systematic nonce reuse that enables complete private key recovery. Analyzing on-chain data reveals that searchers, driven by the need for sub-second response times in sealed-bid auctions, employ predictable nonce patterns. These patterns create linear relationships between signatures, allowing passive attackers to recover private keys using elementary algebra. We provide a compact linear-system formulation for such attacks, including the dangerous case of cross-wallet nonce collisions, and present concrete evidence of exploitable patterns on Polygon. Our findings demonstrate how protocol-induced latency pressures can lead to catastrophic cryptographic failures in production blockchain systems, where a single implementation error compromises multiple accounts simultaneously.
Potential threat/caveat for clean result "The marker is a representational handle, not a behavioural one — sharing it between a villain persona and the assistant transfers no misalignment (HIGH confidence)": this item discusses failure, failures.
- score 100arxiv cs.CR (Cryptography and Security)arxiv:2605.21497unread
Autonomous LLM Agents & CTFs: A Second Look
Youness Bouchari, Matteo Boffa, Marco Mellia, Idilio Drago, Thanh Minh Bui, Dario Rossi · 2026-05-22
arXiv:2605. 21497v1 Announce Type: new Abstract: Large Language Model (LLM) agents are increasingly proposed to automate offensive security tasks, with recent studies reporting near human-level success rates in Capture-the-Flag (CTF) challenges.
Read next because Autonomous LLM Agents & CTFs: A Second Look overlaps with clean result "LoRA persona trained on <A> alone emits <B> at 23.5% when a co-trained partner learns <A>...<B>, vs 0% control on Qwen2.5-7B-Instruct (MODERATE confidence)", clean result "Leakage rate is a usable signal for recovering trigger-shaped phrases on Gaperon-1125-1B without knowing the hidden trigger itself (MODERATE confidence)", clean result "Language-mismatch LoRA SFT on Qwen2.5-7B leaks the trained completion language into bystander directives the model was never trained on, absent under same-language SFT (LOW confidence)". Matching terms: code, strong, class, eval, line, rate, compare, capability. Source: arxiv cs.CR (Cryptography and Security).
arXiv:2605.21497v1 Announce Type: new Abstract: Large Language Model (LLM) agents are increasingly proposed to automate offensive security tasks, with recent studies reporting near human-level success rates in Capture-the-Flag (CTF) challenges. We here revisit these results, providing a second look at these claims. We engineer different agent architectures of increasing complexity and modularity on 30 web-based CTFs challenges spanning 14 vulnerability classes. We instantiate these agents with multiple LLM backbones, and compare them with claude-code, a general-purpose agent that automatically determines its internal architecture. Our evaluation yields three main findings. First, claude-code achieves performance comparable to the engineered architectures (19/30 solved tasks), suggesting that general-purpose agents are strong baselines for offensive security tasks. Second, both our architectures and claude-code struggle in the same challenge categories, revealing persistent barriers that keep current agents below human-level capability. Third, by leveraging our manually designed architectures we can systematically measure the impact of additional components, finding that structured orchestration of specialized roles outperforms monolithic designs, improving run-to-run consistency, and reducing execution costs.
Potential threat/caveat for clean result "LoRA persona trained on <A> alone emits <B> at 23.5% when a co-trained partner learns <A>...<B>, vs 0% control on Qwen2.5-7B-Instruct (MODERATE confidence)": this item discusses evaluation.
- score 100arxiv cs.AI (Artificial Intelligence)arxiv:2605.21458unread
Mind the Sim-to-Real Gap & Think Like a Scientist
Harsh Parikh, Gabriel Levin-Konigsberg, Dominique Perrault-Joncas, Alexander Volfovsky · 2026-05-22
arXiv:2605. 21458v1 Announce Type: new Abstract: Suppose a planner has a pre-trained simulator of a sequential decision problem and the option to run real experiments in the field.
Read next because Mind the Sim-to-Real Gap & Think Like a Scientist overlaps with clean result "Language-mismatch LoRA SFT on Qwen2.5-7B leaks the trained completion language into bystander directives the model was never trained on, absent under same-language SFT (LOW confidence)", clean result "The marker is a representational handle, not a behavioural one — sharing it between a villain persona and the assistant transfers no misalignment (HIGH confidence)", experiment "Follow-up to #354: cascading chunk-binding — does A→B, B→C, C→D propagate the full chain on a recipient trained only to emit A?". Matching terms: under, rate, does, chain, trained, test, lora. Source: arxiv cs.AI (Artificial Intelligence).
arXiv:2605.21458v1 Announce Type: new Abstract: Suppose a planner has a pre-trained simulator of a sequential decision problem and the option to run real experiments in the field. The simulator is cheap to query but inherits confounding and drift from its calibration data. Experimentation is unbiased but consumes one real unit per trial. We study when, and how, the planner should supplement the simulator with experiments. We give three results. First, an extended simulation lemma decomposes the simulator's value error into a calibration--deployment shift that randomization can identify and a parametric residual that no further interaction can reduce. Second, the value gap between the simulator-optimal policy and the optimum splits into a local component, on states the deployed policy already visits, and a reachability component, on states it does not. The reachability component stays bounded away from zero at any horizon under purely passive learning. Third, we propose Fisher-SEP, a simulation-aided experimental policy (SEP) that minimizes the posterior predictive variance of a target policy's value, with reward-only and transition-only specializations. Two case studies illustrate the regimes. In a vending-machine supply chain, front-loaded experimentation overtakes posterior updating once the horizon is long enough to amortize the pilot. In an HIV mobile-testing example with a corridor that separates a well-surveilled region from a poorly-surveilled one, only designed exploration reaches the poorly-surveilled region.
Potential threat/caveat for clean result "Language-mismatch LoRA SFT on Qwen2.5-7B leaks the trained completion language into bystander directives the model was never trained on, absent under same-language SFT (LOW confidence)": this item discusses bias, confound.
- score 100arxiv cs.AI (Artificial Intelligence)arxiv:2605.21413unread
Teaching AI Through Benchmark Construction: QuestBench as a Course-Based Practice for Accountable Knowledge Work
Haiyang Shen, Jiuzheng Wang, Taian Guo, Mugeng Liu, Wenchun Jing, Chongyang Pan, Siqi Zhong, Zhiyang Chen, Weichen Bi, Yudong Han, Xiaoying Bai, Yun Ma · 2026-05-22
arXiv:2605. 21413v2 Announce Type: new Abstract: As AI becomes part of everyday learning, many courses teach students to use it mainly as a productivity tool: how to prompt, search, summarize, write, code, and use tools more efficiently.
Read next because Teaching AI Through Benchmark Construction: QuestBench as a Course-Based Practice for Accountable Knowledge Work overlaps with clean result "LoRA persona trained on <A> alone emits <B> at 23.5% when a co-trained partner learns <A>...<B>, vs 0% control on Qwen2.5-7B-Instruct (MODERATE confidence)", clean result "Leakage rate is a usable signal for recovering trigger-shaped phrases on Gaperon-1125-1B without knowing the hidden trigger itself (MODERATE confidence)", clean result "Language-mismatch LoRA SFT on Qwen2.5-7B leaks the trained completion language into bystander directives the model was never trained on, absent under same-language SFT (LOW confidence)". Matching terms: code, class, rect, under, eval, source, rate, another. Source: arxiv cs.AI (Artificial Intelligence).
arXiv:2605.21413v2 Announce Type: new Abstract: As AI becomes part of everyday learning, many courses teach students to use it mainly as a productivity tool: how to prompt, search, summarize, write, code, and use tools more efficiently. We argue that AI education also needs a setting in which students learn to test AI and understand their own role in judging machine-produced knowledge. To this end, we introduce a course-based practice that teaches AI through benchmark construction, using deep research systems as a concrete example of AI-era knowledge work. Students turn disciplinary knowledge into verifiable expert-level questions, review one another's designs for ambiguity and shortcuts, and evaluate AI systems on the resulting tasks. This activity gives students direct exposure to a powerful tool while asking them to specify what a trustworthy answer would require. The produced benchmark, QuestBench, consists of 256 questions across 14 humanities and social-science domains. Evaluation on QuestBench shows that student-designed tasks reveal hidden failures in current deep research systems: across thirteen evaluated systems, the mean question-level pass rate is only 16.85%, and the best-performing system, GPT-5.5, reaches a 57.58% pass rate. The failures are educationally useful because they show how fluent, source-backed answers can still miss the right query, source, term, or evidence standard. Reflections from five student contributors suggest that benchmark construction can help students see professional knowledge not only as content AI may retrieve, but as the basis for judging AI outputs. We present QuestBench as a benchmark artifact and as a reusable classroom setting for a larger educational question: how students can remain responsible knowledge actors as AI enters learning and professional work. The dataset is available at https://huggingface.co/datasets/PKUAIWeb/QuestBench/tree/main.
Potential threat/caveat for clean result "LoRA persona trained on <A> alone emits <B> at 23.5% when a co-trained partner learns <A>...<B>, vs 0% control on Qwen2.5-7B-Instruct (MODERATE confidence)": this item discusses failure, failures, evaluation, benchmark.
- score 100arxiv cs.AI (Artificial Intelligence)arxiv:2605.21347unread
Insights Generator: Systematic Corpus-Level Trace Diagnostics for LLM Agents
Akshay Manglik (Emily), Apaar Shanker (Emily), Kaustubh Deshpande (Emily), Jason Qin (Emily), Yash Maurya (Emily), Veronica Chatrath (Emily), Vijay S. Kalmath (Emily), Levi Lentz (Emily), Yuan (Emily), Xue · 2026-05-22
arXiv:2605. 21347v2 Announce Type: new Abstract: Diagnosing failures in LLM agents remains largely manual.
Read next because Insights Generator: Systematic Corpus-Level Trace Diagnostics for LLM Agents overlaps with clean result "Only continuous soft prefixes hit both EM axes at once on Qwen-2.5-7B-Instruct: discrete prompt searches split between the alignment objective and the distributional objective, and both discretizations of the soft prefix collapse (MODERATE confidence)", clean result "A pretraining-data-poisoned Qwen3-4B backdoor only fires on the exact trigger tokens — paraphrases don't activate it, and base-model similarity to the trigger doesn't predict which inputs fire (MODERATE confidence)", clean result "The marker is a representational handle, not a behavioural one — sharing it between a villain persona and the assistant transfers no misalignment (HIGH confidence)". Matching terms: eval, token, line, rate, implement, does, test, language. Source: arxiv cs.AI (Artificial Intelligence).
arXiv:2605.21347v2 Announce Type: new Abstract: Diagnosing failures in LLM agents remains largely manual. Practitioners inspect a small subset of execution traces, form ad-hoc hypotheses, and iterate. This process misses patterns that only emerge across trace populations and does not scale to production corpora where individual traces span tens of thousands of tokens. We formalize the problem of corpus-level trace diagnostics. Given a corpus of execution traces, the goal is to produce grounded natural-language insights that characterize systematic behavioral patterns across trace groups, each linked to supporting evidence. We present the Insights Generator (IG), a multi-agent system that answers diagnostic questions by proposing and testing hypotheses across the trace corpus to produce an evidence-backed insights report. We evaluate IG across qualitative and objective dimensions, spanning rubric-based report assessment and downstream performance improvements achieved by implementing IG insights. Human experts using IG reports improve scaffold performance by 30.4pp over the unmodified baseline scaffold, and coding agents leveraging IG-derived insights show consistent and stable gains. Across benchmarks, IG's scout-investigator architecture produces findings comparable in detection coverage to competing approaches, while domain experts rated IG reports as leading depth and evidence quality.
Potential threat/caveat for clean result "Only continuous soft prefixes hit both EM axes at once on Qwen-2.5-7B-Instruct: discrete prompt searches split between the alignment objective and the distributional objective, and both discretizations of the soft prefix collapse (MODERATE confidence)": this item discusses failure, failures, benchmark.
- score 100arxiv cs.AI (Artificial Intelligence)arxiv:2605.21168unread
ScenePilot: Controllable Boundary-Driven Critical Scenario Generation for Autonomous Driving
Qiyu Ruan, Yuxuan Wang, He Li, Zhenning Li, Cheng-zhong Xu · 2026-05-22
arXiv:2605. 21168v1 Announce Type: new Abstract: Safety-critical scenarios are central to evaluating autonomous driving systems, yet their rarity in naturalistic logs makes simulation-based stress testing indispensable.
Read next because ScenePilot: Controllable Boundary-Driven Critical Scenario Generation for Autonomous Driving overlaps with clean result "LoRA persona trained on <A> alone emits <B> at 23.5% when a co-trained partner learns <A>...<B>, vs 0% control on Qwen2.5-7B-Instruct (MODERATE confidence)", clean result "Leakage rate is a usable signal for recovering trigger-shaped phrases on Gaperon-1125-1B without knowing the hidden trigger itself (MODERATE confidence)", clean result "Language-mismatch LoRA SFT on Qwen2.5-7B leaks the trained completion language into bystander directives the model was never trained on, absent under same-language SFT (LOW confidence)". Matching terms: code, eval, line, rate, control, without, trained, capability. Source: arxiv cs.AI (Artificial Intelligence).
arXiv:2605.21168v1 Announce Type: new Abstract: Safety-critical scenarios are central to evaluating autonomous driving systems, yet their rarity in naturalistic logs makes simulation-based stress testing indispensable. Most scenario generation methods treat surrounding agents as adversaries, but they either (i) induce failures without explicitly modeling vehicle-road physical limits, yielding visually extreme yet physically unsolvable crashes, or (ii) enforce physical feasibility or policy feasibility in isolation, which can over-focus on aggressive maneuvers or remain tied to a controller-dependent capability boundary. We propose ScenePilot, a feasibility-guided, boundary-driven framework that targets the boundary band: scenarios that are physically solvable in principle yet still cause the deployed autonomy stack to fail. We formulate generation as constrained multi-objective reinforcement learning, combining an RSS-derived physical-feasibility score $\sigma$ with an online-learned AV-risk predictor $\Phi$, and introduce step-level feasibility-aware shielding to keep exploration near the feasibility boundary while avoiding infeasible artifacts. Experiments on SafeBench with multiple planners show that ScenePilot yields substantially higher collision rates (+6.2 percentage points) while preserving physical validity, and that adversarial fine-tuning on these boundary-band scenarios consistently reduces downstream crash rates. The code is available at https://github.com/QiyuRuan/ScenePilot.
Potential threat/caveat for clean result "LoRA persona trained on <A> alone emits <B> at 23.5% when a co-trained partner learns <A>...<B>, vs 0% control on Qwen2.5-7B-Instruct (MODERATE confidence)": this item discusses failure, failures, adversarial.
- score 100arxiv cs.AI (Artificial Intelligence)arxiv:2605.20911unread
For How Long Should We Be Punching? Learning Action Duration in Fighting Games
Hoang Hai Nguyen, Kurt Driessens, Dennis J. N. J. Soemers · 2026-05-22
arXiv:2605. 20911v1 Announce Type: new Abstract: Fighting games such as Street Fighter II present unique challenges to reinforcement learning (RL) agents due to their fast-paced, real-time nature.
Read next because For How Long Should We Be Punching? Learning Action Duration in Fighting Games overlaps with clean result "LoRA persona trained on <A> alone emits <B> at 23.5% when a co-trained partner learns <A>...<B>, vs 0% control on Qwen2.5-7B-Instruct (MODERATE confidence)", clean result "Leakage rate is a usable signal for recovering trigger-shaped phrases on Gaperon-1125-1B without knowing the hidden trigger itself (MODERATE confidence)", clean result "Language-mismatch LoRA SFT on Qwen2.5-7B leaks the trained completion language into bystander directives the model was never trained on, absent under same-language SFT (LOW confidence)". Matching terms: code, source, rate, implement, compare, does, trained, test. Source: arxiv cs.AI (Artificial Intelligence).
arXiv:2605.20911v1 Announce Type: new Abstract: Fighting games such as Street Fighter II present unique challenges to reinforcement learning (RL) agents due to their fast-paced, real-time nature. In most RL frameworks, agents are hard-coded to make decisions at a fixed interval, typically every frame or every N frames. Although this design ensures timely responses, it restricts the agent's ability to adjust its reaction timing. Acting every frame grants frame-perfect reflexes, which are unrealistic compared to human players, whereas longer fixed intervals reduce computational cost but hinder responsiveness. We consider an alternative decision-making framework in which the agent learns not only what action to take but also for how long to execute it. By jointly predicting both action and duration, the agent can dynamically adapt its responsiveness to different situations in the game. We implement this method using the open-source FightLadder environment with agents trained against scripted built-in bots, systematically testing different frame skip configurations to analyze their influence on performance, responsiveness, and learned behavior. Experiments show that learned timing can match the performance of well-chosen fixed frame skips and encourages repeatable action patterns, but does not ensure robustness on its own. In most cases, we see agents performing best with consistently high frame skip values (i.e., low responsiveness). This strategy makes it easier to learn exploitative strategies where the same action is repeated over and over, which the scripted bots appear to be susceptible to.
Potential threat/caveat for clean result "LoRA persona trained on <A> alone emits <B> at 23.5% when a co-trained partner learns <A>...<B>, vs 0% control on Qwen2.5-7B-Instruct (MODERATE confidence)": this item discusses robustness.
- score 100arxiv cs.AI (Artificial Intelligence)arxiv:2605.20873unread
PlanningBench: Generating Scalable and Verifiable Planning Data for Evaluating and Training Large Language Models
Ziliang Zhao, Zenan Xu, Shuting Wang, Hongjin Qian, Yan Lei, Minda Hu, Zhao Wang, Shihan Dou, Zhicheng Dou, Pluto Zhou · 2026-05-22
arXiv:2605. 20873v1 Announce Type: new Abstract: Planning is a fundamental capability for large language models (LLMs) because such complex tasks require models to coordinate goals, constraints, resources, and long-term consequences into executable and verifiable solutions.
Read next because PlanningBench: Generating Scalable and Verifiable Planning Data for Evaluating and Training Large Language Models overlaps with clean result "Language-mismatch LoRA SFT on Qwen2.5-7B leaks the trained completion language into bystander directives the model was never trained on, absent under same-language SFT (LOW confidence)", clean result "Only continuous soft prefixes hit both EM axes at once on Qwen-2.5-7B-Instruct: discrete prompt searches split between the alignment objective and the distributional objective, and both discretizations of the soft prefix collapse (MODERATE confidence)", clean result "Training one persona to emit a [ZLT] marker without bystanders adopting it has a one-cell-wide LR x epochs window on Qwen2.5-7B-Instruct (LOW confidence)". Matching terms: under, eval, source, line, control, factor, capability, language. Source: arxiv cs.AI (Artificial Intelligence).
arXiv:2605.20873v1 Announce Type: new Abstract: Planning is a fundamental capability for large language models (LLMs) because such complex tasks require models to coordinate goals, constraints, resources, and long-term consequences into executable and verifiable solutions. Existing planning benchmarks, however, usually treat planning data as fixed collections of instances rather than controllable generation targets. This limits scenario coverage, ties difficulty to surface-level proxies rather than structural sources, and offers limited support for scalable generation, automatic verification, or planning-oriented training. We introduce PlanningBench, a framework for generating scalable, diverse, and verifiable planning data for both evaluation and training. PlanningBench starts from real planning scenarios and abstracts practical workflows into a structured taxonomy of more than 30 task types, subtasks, constraint families, and difficulty factors. Guided by this taxonomy, a constraint-driven synthesis pipeline instantiates self-contained planning problems with adaptive difficulty control, quality filtering, and instance-level verification checklists. This shifts planning data construction from fixed benchmark collection to controllable generation while preserving realistic task grounding. We use PlanningBench to evaluate open-source and closed-source frontier LLMs, and find that current models still struggle to produce complete solutions under coupled constraints. Beyond evaluation, reinforcement learning on verified PlanningBench data improves performance on unseen planning benchmarks and broader instruction-following tasks. Further analysis suggests that determinate or well-specified optimal solutions provide clearer reward signals and more stable training dynamics. Overall, PlanningBench provides a controllable source of planning data for diagnosing and improving generalizable planning abilities in LLMs.
Potential threat/caveat for clean result "Language-mismatch LoRA SFT on Qwen2.5-7B leaks the trained completion language into bystander directives the model was never trained on, absent under same-language SFT (LOW confidence)": this item discusses evaluation, benchmark.
- score 100arxiv cs.AI (Artificial Intelligence)arxiv:2605.20834unread
Conditional Equivalence of DPO and RLHF: Implicit Assumption, Failure Modes, and Provable Alignment
Zhiqin Yang, Yonggang Zhang, Wei Xue, Dong Fang, Bo Han, Yike Guo · 2026-05-22
arXiv:2605. 20834v1 Announce Type: new Abstract: Direct Preference Optimization (DPO) has emerged as a popular alternative to Reinforcement Learning from Human Feedback (RLHF), offering theoretical equivalence with simpler implementation.
Read next because Conditional Equivalence of DPO and RLHF: Implicit Assumption, Failure Modes, and Provable Alignment overlaps with clean result "LoRA persona trained on <A> alone emits <B> at 23.5% when a co-trained partner learns <A>...<B>, vs 0% control on Qwen2.5-7B-Instruct (MODERATE confidence)", clean result "Leakage rate is a usable signal for recovering trigger-shaped phrases on Gaperon-1125-1B without knowing the hidden trigger itself (MODERATE confidence)", clean result "Language-mismatch LoRA SFT on Qwen2.5-7B leaks the trained completion language into bystander directives the model was never trained on, absent under same-language SFT (LOW confidence)". Matching terms: code, rect, alignment, soft, rate, implement, trained. Source: arxiv cs.AI (Artificial Intelligence).
arXiv:2605.20834v1 Announce Type: new Abstract: Direct Preference Optimization (DPO) has emerged as a popular alternative to Reinforcement Learning from Human Feedback (RLHF), offering theoretical equivalence with simpler implementation. We prove this equivalence is conditional rather than universal, depending on an implicit assumption frequently violated in practice: the RLHF-optimal policy must prefer human-preferred responses. When this assumption fails, DPO optimizes relative advantage over the reference policy rather than absolute alignment with human preferences, leading to pathological convergence where policies decrease DPO loss while preferring dispreferred responses. We characterize when this assumption is violated, show the existence of an undesirable solution space, and prove that DPO and RLHF optimize fundamentally different objectives in such cases. To address this, we introduce Constrained Preference Optimization (CPO), augmenting RLHF with constraints for provable alignment. We further provide a geometric interpretation through soft margin ranking, revealing that DPO implements margin ranking with potentially negative targets. Our theoretical analysis establishes when DPOs' guarantees hold and provides solutions preserving simplicity with provable alignment. Comprehensive experiments on standard benchmarks demonstrate that CPO achieves state-of-the-art performance. Code is available at: https://github.com/visitworld123/CPO.
Potential threat/caveat for clean result "LoRA persona trained on <A> alone emits <B> at 23.5% when a co-trained partner learns <A>...<B>, vs 0% control on Qwen2.5-7B-Instruct (MODERATE confidence)": this item discusses failure, negative, benchmark.
- score 100arxiv cs.AI (Artificial Intelligence)arxiv:2605.20784unread
Interaction Locality in Hierarchical Recursive Reasoning
Yosuke Miyanishi, Tetsuro Morimura · 2026-05-22
arXiv:2605. 20784v1 Announce Type: new Abstract: Spatial reasoning requires both location-bound computation and location-invariant structure: agents must make local moves while preserving route, object, or constraint-level plans.
Read next because Interaction Locality in Hierarchical Recursive Reasoning overlaps with clean result "LoRA persona trained on <A> alone emits <B> at 23.5% when a co-trained partner learns <A>...<B>, vs 0% control on Qwen2.5-7B-Instruct (MODERATE confidence)", clean result "Leakage rate is a usable signal for recovering trigger-shaped phrases on Gaperon-1125-1B without knowing the hidden trigger itself (MODERATE confidence)", clean result "Language-mismatch LoRA SFT on Qwen2.5-7B leaks the trained completion language into bystander directives the model was never trained on, absent under same-language SFT (LOW confidence)". Matching terms: code, strong, rate, test, model. Source: arxiv cs.AI (Artificial Intelligence).
arXiv:2605.20784v1 Announce Type: new Abstract: Spatial reasoning requires both location-bound computation and location-invariant structure: agents must make local moves while preserving route, object, or constraint-level plans. We propose interaction locality, a task-geometry-aware framework for measuring whether information flow stays within nearby cells or semantic segments, or crosses them. We instantiate the framework with sparse-autoencoder feature ablations and finite-noise activation patching, with structural Jacobian and attention checks reported in the appendix, and apply it to HRM and TRM, two compact hierarchical and recursive reasoning models, on Maze-Hard, Sudoku Extreme, and ARC-AGI. Across these models, activation patching gives the clearest architectural fingerprint: high-level recurrent states tend to write information within nearby cells or same-segment units, while repeated recursive updates accumulate these local writes into broader solution structure. This pattern holds across maze paths, Sudoku constraints, and ARC-AGI object neighborhoods, with the strongest concentration in TRM. To test whether interaction locality extends beyond toy-yet-challenging grid benchmarks, we also apply it to MTU3D, a large-scale embodied 3D scene-grounding model. In this MTU3D setting, causal spatial locality appears primarily at the transition where visual scene features are handed to the downstream grounding module, rather than uniformly throughout the visual encoder. This contrast suggests that the local-to-global handoff observed in HRM and TRM is tied to explicit recursive reasoning dynamics, while embodied 3D models may concentrate causal spatial structure at module boundaries. Interaction locality turns the intuitive local-execution/global-planning story into a reproducible measurement framework for recursive and embodied spatial reasoning.
Potential threat/caveat for clean result "LoRA persona trained on <A> alone emits <B> at 23.5% when a co-trained partner learns <A>...<B>, vs 0% control on Qwen2.5-7B-Instruct (MODERATE confidence)": this item discusses benchmark.
- score 100arxiv cs.AI (Artificial Intelligence)arxiv:2605.20742unread
VBFDD-Agent for Electric Vehicle Battery Fault Detection and Diagnosis: Descriptive Text Modeling of Battery Digital Signals
Joey Chan, Zhen Chen, Ershun Pan · 2026-05-22
arXiv:2605. 20742v1 Announce Type: new Abstract: With the rapid proliferation of electric vehicles, the safety and reliability of lithium-ion batteries have become critical concerns.
Read next because VBFDD-Agent for Electric Vehicle Battery Fault Detection and Diagnosis: Descriptive Text Modeling of Battery Digital Signals overlaps with clean result "LoRA persona trained on <A> alone emits <B> at 23.5% when a co-trained partner learns <A>...<B>, vs 0% control on Qwen2.5-7B-Instruct (MODERATE confidence)", clean result "Leakage rate is a usable signal for recovering trigger-shaped phrases on Gaperon-1125-1B without knowing the hidden trigger itself (MODERATE confidence)", clean result "Language-mismatch LoRA SFT on Qwen2.5-7B leaks the trained completion language into bystander directives the model was never trained on, absent under same-language SFT (LOW confidence)". Matching terms: strong, text, eval, source, rate, language, model. Source: arxiv cs.AI (Artificial Intelligence).
arXiv:2605.20742v1 Announce Type: new Abstract: With the rapid proliferation of electric vehicles, the safety and reliability of lithium-ion batteries have become critical concerns. Effective anomaly detection is essential for ensuring safe battery operation. However, as battery systems and operating scenarios become increasingly complex, battery fault diagnosis and maintenance require stronger cross-domain adaptability and human-AI collaboration. Traditional fault detection and diagnosis methods are usually designed for specific scenarios and predefined workflows, making them less effective in complex real-world applications. To address the scarcity of open-source battery fault report corpora and the lack of unified maintenance knowledge representation, this study proposes a descriptive text modeling approach for battery signal reports. Monitoring signals, statistical features, anomaly records, and state assessment results are transformed into structured and readable natural language descriptions, forming a language corpus for battery health diagnosis and maintenance. Based on this corpus, we propose VBFDD-Agent, a vehicle battery fault detection and diagnosis agent for automotive-grade battery systems. VBFDD-Agent integrates descriptive battery-state texts, historical case retrieval, local maintenance manuals, and large language model reasoning to generate structured diagnostic results and maintenance recommendations. Experiments show that the proposed framework can accurately perform anomaly monitoring based on descriptive textual representations and provide flexible, efficient, and actionable maintenance suggestions. Expert evaluation further confirms the practical value of the generated recommendations. Overall, VBFDD-Agent extends traditional battery diagnosis from label prediction to interpretable and maintenance-oriented decision support.
Potential threat/caveat for clean result "LoRA persona trained on <A> alone emits <B> at 23.5% when a co-trained partner learns <A>...<B>, vs 0% control on Qwen2.5-7B-Instruct (MODERATE confidence)": this item discusses evaluation.
- score 100arxiv cs.AI (Artificial Intelligence)arxiv:2605.20690unread
Declarative Data Services: Structured Agentic Discovery for Composing Data Systems
Shanshan Ye, Duo Lu · 2026-05-22
arXiv:2605. 20690v1 Announce Type: new Abstract: Agentic discovery has shown that LLM-driven search can find novel algorithms, designs, and code under benchmark conditions.
Read next because Declarative Data Services: Structured Agentic Discovery for Composing Data Systems overlaps with clean result "LoRA persona trained on <A> alone emits <B> at 23.5% when a co-trained partner learns <A>...<B>, vs 0% control on Qwen2.5-7B-Instruct (MODERATE confidence)", clean result "Leakage rate is a usable signal for recovering trigger-shaped phrases on Gaperon-1125-1B without knowing the hidden trigger itself (MODERATE confidence)", clean result "Language-mismatch LoRA SFT on Qwen2.5-7B leaks the trained completion language into bystander directives the model was never trained on, absent under same-language SFT (LOW confidence)". Matching terms: code, latin, under, line, does, position. Source: arxiv cs.AI (Artificial Intelligence).
arXiv:2605.20690v1 Announce Type: new Abstract: Agentic discovery has shown that LLM-driven search can find novel algorithms, designs, and code under benchmark conditions. Translating the paradigm to multi-system data backends surfaces a harder problem: the search space is heterogeneous, the verifier is whether a deployed stack actually runs, and composition knowledge is unevenly captured in pretraining. Unbounded agentic discovery, a coding agent iterating on failure-log feedback, fails to converge consistently on a working stack even when iteration and explicit composition knowledge are added. We propose Declarative Data Services (DDS), an architecture for structured agentic discovery of data-system compositions from declarative user intent. The framework owns four typed contracts at successive layers (intent, operator DAG, per-system skills, runtime attribution) that decompose the global search into bounded sub-searches; sub-agents search each typed space, while the framework provides the channels by which knowledge flows forward as inline skill citations and errors route backward as typed signals. As a proof of life on a trading-backend workload, DDS converges where unbounded discovery does not; runtime failures become skill patches that the next deployment cites inline. We position this as an early prototype reporting lessons from real-world data-system composition.
Potential threat/caveat for clean result "LoRA persona trained on <A> alone emits <B> at 23.5% when a co-trained partner learns <A>...<B>, vs 0% control on Qwen2.5-7B-Instruct (MODERATE confidence)": this item discusses failure, failures, benchmark.
- score 100arxiv cs.AI (Artificial Intelligence)arxiv:2605.20630unread
Evaluating Temporal Semantic Caching and Workflow Optimization in Agentic Plan-Execute Pipelines
Alimurtaza Mustafa Merchant, Krish Veera, Sajal Kumar Goyla, Shambhawi Bhure, Dhaval Patel, Kaoutar El Maghraoui · 2026-05-22
arXiv:2605. 20630v1 Announce Type: new Abstract: Industrial asset operations workflows are latency-sensitive because a single user query may require coordination over sensor data, work orders, failure modes, forecasting tools, and domain-specific agents.
Read next because Evaluating Temporal Semantic Caching and Workflow Optimization in Agentic Plan-Execute Pipelines overlaps with clean result "Language-mismatch LoRA SFT on Qwen2.5-7B leaks the trained completion language into bystander directives the model was never trained on, absent under same-language SFT (LOW confidence)", clean result "Coupling evil personas with wrong answers fails to protect Qwen2.5-7B from EM-induced alignment collapse — and the apparent capability ordering across coupling conditions is mostly eval contamination (LOW confidence)", clean result "Only continuous soft prefixes hit both EM axes at once on Qwen-2.5-7B-Instruct: discrete prompt searches split between the alignment objective and the distributional objective, and both discretizations of the soft prefix collapse (MODERATE confidence)". Matching terms: rect, correct, eval, line. Source: arxiv cs.AI (Artificial Intelligence).
arXiv:2605.20630v1 Announce Type: new Abstract: Industrial asset operations workflows are latency-sensitive because a single user query may require coordination over sensor data, work orders, failure modes, forecasting tools, and domain-specific agents. We evaluate this problem on AssetOpsBench (AOB), an industrial agent benchmark whose plan-execute pipeline exposes repeated overhead from tool discovery, LLM planning, MCP tool execution, and final summarization. Existing LLM caching techniques such as KV-cache reuse and embedding-based semantic caching were designed for chatbot serving and break down when output validity depends on time, asset, or sensor parameters. We propose two complementary optimization layers for AOB plan-execute pipelines: a temporal semantic cache and a set of MCP workflow optimizations combining disk-backed tool-discovery caching and dependency-aware parallel step execution. MCP workflow optimizations corresponded to a 1.67x speedup and reduced median end-to-end latency by about 40.0% while the temporal-cache benchmark achieved a median of 30.6x speedup on cache hits. Beyond the speedup, our results expose a concrete failure mode of pure semantic caching for parameter-rich industrial queries, providing a critical analysis of how caching choices interact with evaluation correctness in MCP-backed agent benchmarks.
Potential threat/caveat for clean result "Language-mismatch LoRA SFT on Qwen2.5-7B leaks the trained completion language into bystander directives the model was never trained on, absent under same-language SFT (LOW confidence)": this item discusses failure, evaluation, benchmark.
- score 100arxiv cs.AI (Artificial Intelligence)arxiv:2605.20618unread
COAgents: Multi-Agent Framework to Learn and Navigate Routing Problems Search Space
Oleksandr Yakovenko, Mahdi Mostajabdaveh, Cheikh Ahmed, Abdullah Ali Sivas, Xiaorui Li, Zirui Zhou, Mao Kun · 2026-05-22
arXiv:2605. 20618v1 Announce Type: new Abstract: Although Vehicle Routing Problems (VRP) are essential to many real-world systems, they remain computationally intractable at scale due to their combinatorial complexity.
Read next because COAgents: Multi-Agent Framework to Learn and Navigate Routing Problems Search Space overlaps with clean result "LoRA persona trained on <A> alone emits <B> at 23.5% when a co-trained partner learns <A>...<B>, vs 0% control on Qwen2.5-7B-Instruct (MODERATE confidence)", clean result "Leakage rate is a usable signal for recovering trigger-shaped phrases on Gaperon-1125-1B without knowing the hidden trigger itself (MODERATE confidence)", clean result "Language-mismatch LoRA SFT on Qwen2.5-7B leaks the trained completion language into bystander directives the model was never trained on, absent under same-language SFT (LOW confidence)". Matching terms: code, strong, text, line, rate, control, lora, model. Source: arxiv cs.AI (Artificial Intelligence).
arXiv:2605.20618v1 Announce Type: new Abstract: Although Vehicle Routing Problems (VRP) are essential to many real-world systems, they remain computationally intractable at scale due to their combinatorial complexity. Traditional heuristics rely on handcrafted rules for local improvements and occasional \textit{jumps} to escape local minima, but often struggle to generalize across diverse instances. We introduce \textbf{COAgents}, a cooperative multi-agent framework that models the search process as a graph: nodes represent solutions, and edges correspond to either local refinements or large perturbations for diversification (i.e., jumps). A \textit{Partial Search Graph} (PSG) is dynamically constructed during search, enabling COAgents to train a Node Selection Agent and a Move Selection Agent to guide intensification, and a Jump Agent to trigger well-timed explorations of new regions. Unlike end-to-end learning approaches, COAgents cleanly separates problem-agnostic search control from compact domain-specific encoding, facilitating adaptability across tasks. Extensive experiments on the CVRP and VRPTW benchmarks show that COAgents remains competitive with several learn-to-search baselines on CVRP and sets a new state of the art among learning-based methods on the more challenging VRPTW instances, reducing the gap to the best-known solutions by 14\% at $N\!=\!100$ and 44\% at $N\!=\!50$ relative to the strongest neural solver (POMO), and by 21\% and 40\% respectively relative to ALNS. Code is available at https://github.com/mahdims/COAgents.
Potential threat/caveat for clean result "LoRA persona trained on <A> alone emits <B> at 23.5% when a co-trained partner learns <A>...<B>, vs 0% control on Qwen2.5-7B-Instruct (MODERATE confidence)": this item discusses benchmark.
- score 100arxiv cs.AI (Artificial Intelligence)arxiv:2605.20554unread
Personality Engineering with AI Agents: A New Methodology for Negotiation Research
Michelle A. Vaccaro, Jared R. Curhan · 2026-05-22
arXiv:2605. 20554v1 Announce Type: new Abstract: According to canonical negotiation theory, people's success in a negotiation depends on how well they balance competing demands--empathizing and asserting, demonstrating concern for other and concern for self, being soft on the people and hard on the problem.
Read next because Personality Engineering with AI Agents: A New Methodology for Negotiation Research overlaps with clean result "LoRA persona trained on <A> alone emits <B> at 23.5% when a co-trained partner learns <A>...<B>, vs 0% control on Qwen2.5-7B-Instruct (MODERATE confidence)", clean result "Leakage rate is a usable signal for recovering trigger-shaped phrases on Gaperon-1125-1B without knowing the hidden trigger itself (MODERATE confidence)", clean result "Language-mismatch LoRA SFT on Qwen2.5-7B leaks the trained completion language into bystander directives the model was never trained on, absent under same-language SFT (LOW confidence)". Matching terms: persona, class, under, soft, eval, control, test. Source: arxiv cs.AI (Artificial Intelligence).
arXiv:2605.20554v1 Announce Type: new Abstract: According to canonical negotiation theory, people's success in a negotiation depends on how well they balance competing demands--empathizing and asserting, demonstrating concern for other and concern for self, being soft on the people and hard on the problem. Yet people struggle to manage these tensions, so researchers have lacked the ability to rigorously test the field's prescriptions under controlled conditions. AI agents do not face the same limitations, and their precision, repertoire, consistency, and scalability enable a new class of experiments to contribute to negotiation theory. In this article, we introduce personality engineering: a methodology that uses AI agents to precisely parameterize, manipulate, and evaluate negotiator personality. We propose using the interpersonal circumplex--and its two core dimensions of warmth and dominance--as a foundational coordinate system for the field. This approach offers both a rigorous methodology for testing classic negotiation theories and a practical guide for designing the personalities of AI negotiation agents.
Potential threat/caveat for clean result "LoRA persona trained on <A> alone emits <B> at 23.5% when a co-trained partner learns <A>...<B>, vs 0% control on Qwen2.5-7B-Instruct (MODERATE confidence)": this item discusses limitation, limitations.
- score 100arxiv cs.AI (Artificial Intelligence)arxiv:2605.20530unread
AgentAtlas: Beyond Outcome Leaderboards for LLM Agents
Parsa Mazaheri, Kasra Mazaheri · 2026-05-22
arXiv:2605. 20530v1 Announce Type: new Abstract: Large language model agents now act on codebases, browsers, operating systems, calendars, files, and tool ecosystems, but the benchmarks used to evaluate them are fragmented: each emphasizes a different unit of measurement (final task success, tool-call validity, repeated-pass consistency, trajectory safety, or attack robustness).
Read next because AgentAtlas: Beyond Outcome Leaderboards for LLM Agents overlaps with clean result "LoRA persona trained on <A> alone emits <B> at 23.5% when a co-trained partner learns <A>...<B>, vs 0% control on Qwen2.5-7B-Instruct (MODERATE confidence)", clean result "Leakage rate is a usable signal for recovering trigger-shaped phrases on Gaperon-1125-1B without knowing the hidden trigger itself (MODERATE confidence)", clean result "Language-mismatch LoRA SFT on Qwen2.5-7B leaks the trained completion language into bystander directives the model was never trained on, absent under same-language SFT (LOW confidence)". Matching terms: code, text, under, eval, source, line, rate, control. Source: arxiv cs.AI (Artificial Intelligence).
arXiv:2605.20530v1 Announce Type: new Abstract: Large language model agents now act on codebases, browsers, operating systems, calendars, files, and tool ecosystems, but the benchmarks used to evaluate them are fragmented: each emphasizes a different unit of measurement (final task success, tool-call validity, repeated-pass consistency, trajectory safety, or attack robustness). A line of 2024-2025 work has converged on the diagnosis that a single accuracy column is no longer the right unit of comparison for deployable agents. AgentAtlas extends this line of work with four components: (i) a six-state control-decision taxonomy (Act / Ask / Refuse / Stop / Confirm / Recover); (ii) a nine-category trajectory-failure taxonomy with two orthogonal hierarchical labels (primary_error_source, impact); (iii) a taxonomy-aware vs. taxonomy-blind methodology that measures how much of a model's apparent capability comes from the supervision in the prompt; and (iv) a benchmark-coverage audit mapping fifteen agent benchmarks against six behavioral axes. To demonstrate the methodology we run a small fixed eight-model set (1,342 generated items, four frontier closed and four open-weight) under both prompt modes. Removing the explicit label menu drops every model's trajectory accuracy by 14-40 pp to a tight 0.54-0.62 floor regardless of family, and no single model wins on all three of control accuracy, trajectory diagnosis, and tool-context utility retention. We treat the synthetic run as a measurement-protocol demonstration, not a benchmark release.
Potential threat/caveat for clean result "LoRA persona trained on <A> alone emits <B> at 23.5% when a co-trained partner learns <A>...<B>, vs 0% control on Qwen2.5-7B-Instruct (MODERATE confidence)": this item discusses failure, robustness, benchmark.
- score 100arxiv cs.AI (Artificial Intelligence)arxiv:2605.20520unread
Open-World Evaluations for Measuring Frontier AI Capabilities
Sayash Kapoor, Peter Kirgis, Andrew Schwartz, Stephan Rabanser, J. J. Allaire, Rishi Bommasani, Harry Coppock, Magda Dubois, Gillian K Hadfield, Andrew B. Hall, Sara Hooker, Seth Lazar, Steve Newman, Dimitris Papailiopoulos, Shoshannah Tekofsky, Helen Toner, Cozmin Ududec, Arvind Narayanan · 2026-05-22
arXiv:2605. 20520v1 Announce Type: new Abstract: Benchmark-based evaluation remains important for tracking frontier AI progress.
Read next because Open-World Evaluations for Measuring Frontier AI Capabilities overlaps with clean result "Leakage rate is a usable signal for recovering trigger-shaped phrases on Gaperon-1125-1B without knowing the hidden trigger itself (MODERATE confidence)", clean result "Language-mismatch LoRA SFT on Qwen2.5-7B leaks the trained completion language into bystander directives the model was never trained on, absent under same-language SFT (LOW confidence)", clean result "Only continuous soft prefixes hit both EM axes at once on Qwen-2.5-7B-Instruct: discrete prompt searches split between the alignment objective and the distributional objective, and both discretizations of the soft prefix collapse (MODERATE confidence)". Matching terms: class, under, eval, project, capability. Source: arxiv cs.AI (Artificial Intelligence).
arXiv:2605.20520v1 Announce Type: new Abstract: Benchmark-based evaluation remains important for tracking frontier AI progress. But it can both overstate and understate deployed capability because it privileges tasks that can be precisely specified, automatically graded, easy to optimize for, and run with low budgets and short time horizons. We advocate for a complementary class of evaluations, which we term open-world evaluations: long-horizon, messy, real-world tasks assessed through small-sample qualitative analysis rather than benchmark-scale automation. In this paper we survey recent open-world evaluations, identify their strengths and limitations, and introduce CRUX (Collaborative Research for Updating AI eXpectations), a project for conducting such evaluations regularly. As a first instance, we task an AI agent with developing and publishing a simple iOS application to the Apple App Store. The agent completed the task with only a single avoidable manual intervention, suggesting that open-world evaluations can provide early warning of capabilities that may soon become widespread. We conclude with recommendations for designing and reporting open-world evals.
Potential threat/caveat for clean result "Leakage rate is a usable signal for recovering trigger-shaped phrases on Gaperon-1125-1B without knowing the hidden trigger itself (MODERATE confidence)": this item discusses limitation, limitations, evaluation, benchmark.
- score 100arxiv cs.AI (Artificial Intelligence)arxiv:2605.20490unread
$ECUAS_n$: A family of metrics for principled evaluation of uncertainty-augmented systems
Lautaro Estienne, Erik Ernst, Mat\'ias Vera, Pablo Piantanida, Luciana Ferrer · 2026-05-22
arXiv:2605. 20490v2 Announce Type: new Abstract: In high-stakes automated decision-making, access to predictive uncertainty is essential for enabling users -- human or downstream systems -- to accept or reject predictions based on application-specific cost trade-offs.
Read next because $ECUAS_n$: A family of metrics for principled evaluation of uncertainty-augmented systems overlaps with clean result "Leakage rate is a usable signal for recovering trigger-shaped phrases on Gaperon-1125-1B without knowing the hidden trigger itself (MODERATE confidence)", clean result "Language-mismatch LoRA SFT on Qwen2.5-7B leaks the trained completion language into bystander directives the model was never trained on, absent under same-language SFT (LOW confidence)", clean result "Coupling evil personas with wrong answers fails to protect Qwen2.5-7B from EM-induced alignment collapse — and the apparent capability ordering across coupling conditions is mostly eval contamination (LOW confidence)". Matching terms: class, rect, under, correct, eval, rate, control. Source: arxiv cs.AI (Artificial Intelligence).
arXiv:2605.20490v2 Announce Type: new Abstract: In high-stakes automated decision-making, access to predictive uncertainty is essential for enabling users -- human or downstream systems -- to accept or reject predictions based on application-specific cost trade-offs. Such uncertainty-augmented (UA) systems -- i.e., systems that output both predictions and uncertainty scores -- are currently being assessed in the literature in a variety of ways, using separate metrics to evaluate the predictions and the uncertainty scores, setting a cost function with a fixed rejection cost or integrating over a coverage-risk curve. We argue that these evaluation approaches are inadequate for assessing overall performance of the UA system for decision making under uncertainty and propose a novel family of metrics, $ECUAS_n$, formulated as proper scoring rules for the task of interest. The parameter $n$ controls the trade-off between the cost of incorrect predictions and imperfect uncertainties depending on the needs of the use-case. We demonstrate the advantages of the $ECUAS_n$ metrics both theoretically and empirically, through experiments on diverse classification and generation datasets, including a manually annotated subset of TriviaQA.
Potential threat/caveat for clean result "Leakage rate is a usable signal for recovering trigger-shaped phrases on Gaperon-1125-1B without knowing the hidden trigger itself (MODERATE confidence)": this item discusses evaluation.
- score 100arxiv cs.AI (Artificial Intelligence)arxiv:2605.20467unread
High Quality Embeddings for Horn Logic Reasoning
Yifan Zhang, Yasir White, Dean Clark, Joseph Sanchez, Jevon Lipsey, Ashely Hirst, Jeff Heflin · 2026-05-22
arXiv:2605. 20467v1 Announce Type: new Abstract: Neural networks can be trained to rank the choices made by logical reasoners, resulting in more efficient searches for answers.
Read next because High Quality Embeddings for Horn Logic Reasoning overlaps with clean result "Coupling evil personas with wrong answers fails to protect Qwen2.5-7B from EM-induced alignment collapse — and the apparent capability ordering across coupling conditions is mostly eval contamination (LOW confidence)", clean result "Only continuous soft prefixes hit both EM axes at once on Qwen-2.5-7B-Instruct: discrete prompt searches split between the alignment objective and the distributional objective, and both discretizations of the soft prefix collapse (MODERATE confidence)", experiment "Follow-up to #354: cascading chunk-binding — does A→B, B→C, C→D propagate the full chain on a recipient trained only to emit A?". Matching terms: good, eval, trained. Source: arxiv cs.AI (Artificial Intelligence).
arXiv:2605.20467v1 Announce Type: new Abstract: Neural networks can be trained to rank the choices made by logical reasoners, resulting in more efficient searches for answers. A key step in this process is creating useful embeddings, i.e., numeric representations of logical statements. This paper introduces and evaluates several approaches to creating embeddings that result in better downstream results. We train embeddings using triplet loss, which requires examples consisting of an anchor, a positive example, and a negative example. We introduce three ideas: generating anchors that are more likely to have repeated terms, generating positive and negative examples in a way that ensures a good balance between easy, medium, and hard examples, and periodically emphasizing the hardest examples during training. We conduct several experiments to evaluate this approach, including a comparison of different embeddings across different knowledge bases, in an attempt to identify what characteristics make an embedding well-suited to a particular reasoning task.
Potential threat/caveat for clean result "Coupling evil personas with wrong answers fails to protect Qwen2.5-7B from EM-induced alignment collapse — and the apparent capability ordering across coupling conditions is mostly eval contamination (LOW confidence)": this item discusses negative.
- score 100arxiv cs.AI (Artificial Intelligence)arxiv:2605.20425unread
AgentCo-op: Retrieval-Based Synthesis of Interoperable Multi-Agent Workflows
Shuaike Shen, Wenduo Cheng, Shike Wang, Mingqian Ma, Jian Ma · 2026-05-22
arXiv:2605. 20425v1 Announce Type: new Abstract: Designing multi-agent workflows is especially difficult in open-ended scientific settings where tasks lack curated training sets, reliable scalar evaluation metrics, and standardized interfaces between existing tools and agents.
Read next because AgentCo-op: Retrieval-Based Synthesis of Interoperable Multi-Agent Workflows overlaps with clean result "LoRA persona trained on <A> alone emits <B> at 23.5% when a co-trained partner learns <A>...<B>, vs 0% control on Qwen2.5-7B-Instruct (MODERATE confidence)", clean result "Language-mismatch LoRA SFT on Qwen2.5-7B leaks the trained completion language into bystander directives the model was never trained on, absent under same-language SFT (LOW confidence)", clean result "Only continuous soft prefixes hit both EM axes at once on Qwen-2.5-7B-Instruct: discrete prompt searches split between the alignment objective and the distributional objective, and both discretizations of the soft prefix collapse (MODERATE confidence)". Matching terms: marker, under, eval, line, rate, without. Source: arxiv cs.AI (Artificial Intelligence).
arXiv:2605.20425v1 Announce Type: new Abstract: Designing multi-agent workflows is especially difficult in open-ended scientific settings where tasks lack curated training sets, reliable scalar evaluation metrics, and standardized interfaces between existing tools and agents. We propose AgentCo-op, a retrieval-based synthesis framework that composes reusable skills, tools, and external agents into executable workflows through typed artifact handoffs, then applies bounded self-guided local repair to implicated components when execution evidence indicates failure. In two open-world genomics case studies, AgentCo-op composes independently developed scientific agents and external tool repositories into auditable workflows without redesigning them or running global topology search. It coordinates specialized agents for spatial transcriptomics and gene-set interpretation to enable collaborative discovery from spatial transcriptomics data, and builds a parallel workflow for cross-modality marker analysis on single-cell multiome data. AgentCo-op can also import a searched workflow as a structural prior and improve it by grounding nodes with retrieved components and applying local repair, showing that synthesis and search are complementary. On six coding, math, and question-answering benchmarks, AgentCo-op achieves the best result on four benchmarks and the best average score under a unified backbone setting, while consistently reducing per-task cost relative to multi-agent baselines. Together, these results suggest that retrieval-based synthesis can extend automated agentic workflow design beyond benchmark-optimized agent graphs to open-world workflows built from existing agents, tools, and typed artifacts.
Potential threat/caveat for clean result "LoRA persona trained on <A> alone emits <B> at 23.5% when a co-trained partner learns <A>...<B>, vs 0% control on Qwen2.5-7B-Instruct (MODERATE confidence)": this item discusses failure, evaluation, benchmark.
- score 100arxiv cs.AI (Artificial Intelligence)arxiv:2605.20423unread
OSCToM: RL-Guided Adversarial Generation for High-Order Theory of Mind
Sharmin Sultana Srishty, Kazi Mahathir Rahman, Malaika Parizat Sakkhi, Samia Shahid Prianna, Shaikhul Islam Sinat · 2026-05-22
arXiv:2605. 20423v1 Announce Type: new Abstract: Large Language Models (LLMs) perform well on many language tasks, but their Theory of Mind (ToM) reasoning is still uneven in complex social settings.
Read next because OSCToM: RL-Guided Adversarial Generation for High-Order Theory of Mind overlaps with clean result "LoRA persona trained on <A> alone emits <B> at 23.5% when a co-trained partner learns <A>...<B>, vs 0% control on Qwen2.5-7B-Instruct (MODERATE confidence)", clean result "Leakage rate is a usable signal for recovering trigger-shaped phrases on Gaperon-1125-1B without knowing the hidden trigger itself (MODERATE confidence)", clean result "Language-mismatch LoRA SFT on Qwen2.5-7B leaks the trained completion language into bystander directives the model was never trained on, absent under same-language SFT (LOW confidence)". Matching terms: code, rate, compare, project, position, another, test, language. Source: arxiv cs.AI (Artificial Intelligence).
arXiv:2605.20423v1 Announce Type: new Abstract: Large Language Models (LLMs) perform well on many language tasks, but their Theory of Mind (ToM) reasoning is still uneven in complex social settings. Existing benchmarks, including ExploreToM, do not always test the recursive beliefs and information asymmetries that make these settings difficult. This paper presents OSCToM (Observer-Self Conflict Theory of Mind), an approach for modeling nested belief conflicts in LLM-based ToM tasks. The key case is one in which an observer's view of another agent conflicts with the observer's own belief state. Such cases go beyond simple perspective-taking and require recursive, multi-layered reasoning. OSCToM combines reinforcement learning (RL), an extended domain-specific language, and compositional surrogate models to generate observer-self conflicts. In our experiments, OSCToM-8B gives the best overall result among the systems tested. It improves on the reported ExploreToM results on FANToM and remains competitive on Hi-ToM and BigToM. On the information-asymmetric FANToM benchmark, OSCToM reaches 76% accuracy, compared with the 0.2% reported by ExploreToM. The data-synthesis procedure is also 6x more efficient, indicating that targeted training data can help smaller models handle advanced cognitive reasoning. The project code is available at https://github.com/sharminsrishty/osct.
Potential threat/caveat for clean result "LoRA persona trained on <A> alone emits <B> at 23.5% when a co-trained partner learns <A>...<B>, vs 0% control on Qwen2.5-7B-Instruct (MODERATE confidence)": this item discusses adversarial, benchmark.
- score 100arxiv cs.AI (Artificial Intelligence)arxiv:2605.20190unread
Tool-Augmented Agent for Closed-loop Optimization,Simulation,and Modeling Orchestration
Liyuan Deng, Shujian Deng, Yongkang Chen, Yongkang Dai, Zhihang Zhong, Linyang Li, Xiao Sun, Yilei Shi, Huaxi Huang · 2026-05-22
arXiv:2605. 20190v1 Announce Type: new Abstract: Iterative industrial design-simulation optimization is bottlenecked by the CAD-CAE semantic gap: translating simulation feedback into valid geometric edits under diverse, coupled constraints.
Read next because Tool-Augmented Agent for Closed-loop Optimization,Simulation,and Modeling Orchestration overlaps with clean result "LoRA persona trained on <A> alone emits <B> at 23.5% when a co-trained partner learns <A>...<B>, vs 0% control on Qwen2.5-7B-Instruct (MODERATE confidence)", clean result "Leakage rate is a usable signal for recovering trigger-shaped phrases on Gaperon-1125-1B without knowing the hidden trigger itself (MODERATE confidence)", clean result "Language-mismatch LoRA SFT on Qwen2.5-7B leaks the trained completion language into bystander directives the model was never trained on, absent under same-language SFT (LOW confidence)". Matching terms: strong, fill, latin, under, eval, source, rate, chain. Source: arxiv cs.AI (Artificial Intelligence).
arXiv:2605.20190v1 Announce Type: new Abstract: Iterative industrial design-simulation optimization is bottlenecked by the CAD-CAE semantic gap: translating simulation feedback into valid geometric edits under diverse, coupled constraints. To fill this gap, we propose COSMO-Agent (Closed-loop Optimization, Simulation, and Modeling Orchestration), a tool-augmented reinforcement learning (RL) framework that teaches LLMs to complete the closed-loop CAD-CAE process. Specifically, we cast CAD generation, CAE solving, result parsing, and geometry revision as an interactive RL environment, where an LLM learns to orchestrate external tools and revise parametric geometries until constraints are satisfied. To make this learning stable and industrially usable, we design a multi-constraint reward that jointly encourages feasibility, toolchain robustness, and structured output validity. In addition, we contribute an industry-aligned dataset that covers 25 component categories with executable CAD-CAE tasks to support realistic training and evaluation. Experiments show that COSMO-Agent training substantially improves small open-source LLMs for constraint-driven design, exceeding large open-source and strong closed-source models in feasibility, efficiency, and stability.
Potential threat/caveat for clean result "LoRA persona trained on <A> alone emits <B> at 23.5% when a co-trained partner learns <A>...<B>, vs 0% control on Qwen2.5-7B-Instruct (MODERATE confidence)": this item discusses robustness, evaluation.
- score 100arxiv cs.AI (Artificial Intelligence)arxiv:2605.20189unread
SOLAR: A Self-Optimizing Open-Ended Autonomous Agent for Lifelong Learning and Continual Adaptation
Nitin Vetcha, Dianbo Liu · 2026-05-22
arXiv:2605. 20189v1 Announce Type: new Abstract: Despite the remarkable success of large language models (LLMs), they still face bottlenecks while deploying in dynamic, real-world settings with primary challenges being concept drift and the high cost of gradient-based adaptation.
Read next because SOLAR: A Self-Optimizing Open-Ended Autonomous Agent for Lifelong Learning and Continual Adaptation overlaps with clean result "LoRA persona trained on <A> alone emits <B> at 23.5% when a co-trained partner learns <A>...<B>, vs 0% control on Qwen2.5-7B-Instruct (MODERATE confidence)", clean result "Leakage rate is a usable signal for recovering trigger-shaped phrases on Gaperon-1125-1B without knowing the hidden trigger itself (MODERATE confidence)", clean result "Language-mismatch LoRA SFT on Qwen2.5-7B leaks the trained completion language into bystander directives the model was never trained on, absent under same-language SFT (LOW confidence)". Matching terms: strong, line, rate, without, test, lora, language, model. Source: arxiv cs.AI (Artificial Intelligence).
arXiv:2605.20189v1 Announce Type: new Abstract: Despite the remarkable success of large language models (LLMs), they still face bottlenecks while deploying in dynamic, real-world settings with primary challenges being concept drift and the high cost of gradient-based adaptation. Traditional fine-tuning (FT) struggles to adapt to non-stationary data streams without resulting in catastrophic for getting or requiring extensive manual data curation. To address these limitations within the streaming and continual learning paradigm, we propose the Self-Optimizing Lifelong Autonomous Reasoner (SOLAR) which is an open-ended autonomous agent that leverages parameter-level meta-learning to self-improve, treating model weights as an environment for exploration. It initiates the process by consolidating a strong prior over common-sense knowledge making it effective for transfer-learning. By utilizing a multi-level reinforcement learning approach, SOLAR autonomously discovers adaptation strategies, enabling efficient test-time adaptation to unseen domains. Crucially, SOLAR maintains an evolving knowledge base of valid modification strategies, implicitly acting as an episodic memory buffer to balance plasticity (adaptation to new tasks) and stability (retention of meta-knowledge). Experiments demonstrate that SOLAR outperforms strong baselines on common-sense, mathematical, medical, coding, social and logical reasoning tasks, marking a significant step toward autonomous agents capable of lifelong adaptation in evolving environments.
Potential threat/caveat for clean result "LoRA persona trained on <A> alone emits <B> at 23.5% when a co-trained partner learns <A>...<B>, vs 0% control on Qwen2.5-7B-Instruct (MODERATE confidence)": this item discusses limitation, limitations.
- score 100arxiv cs.CL (NLP)arxiv:2605.22202unread
Structure Retention in Embedding Spaces as a Predictor of Benchmark Performance
Amanda Myntti, Jenna Kanerva, Veronika Laippala, Filip Ginter · 2026-05-22
arXiv:2605. 22202v1 Announce Type: new Abstract: In this paper, we show that high-performing embedding models organize their embedding spaces in a consistent way.
Read next because Structure Retention in Embedding Spaces as a Predictor of Benchmark Performance overlaps with clean result "LoRA persona trained on <A> alone emits <B> at 23.5% when a co-trained partner learns <A>...<B>, vs 0% control on Qwen2.5-7B-Instruct (MODERATE confidence)", clean result "Leakage rate is a usable signal for recovering trigger-shaped phrases on Gaperon-1125-1B without knowing the hidden trigger itself (MODERATE confidence)", clean result "Language-mismatch LoRA SFT on Qwen2.5-7B leaks the trained completion language into bystander directives the model was never trained on, absent under same-language SFT (LOW confidence)". Matching terms: strong, text, class, under, eval, line, model. Source: arxiv cs.CL (NLP).
arXiv:2605.22202v1 Announce Type: new Abstract: In this paper, we show that high-performing embedding models organize their embedding spaces in a consistent way. We evaluate 25 contemporary embedding models on five MTEB tasks spanning four diverse task categories (retrieval, bitext mining, pair classification, and summarization) in both English and multilingual settings, and reveal that nearest-neighbor overlap and magnitude differences in independent component analysis (ICA) between paired text instances strongly correlate (even up to 0.97) with performance on the given task. Ultimately, we show that embedding tasks display varying degrees of linearity and reliance on retention of local information. Our results further the understanding of embeddings, their relation to model performance, and shed light on possible future training objectives and optimizing conditional embeddings.
Potential threat/caveat for clean result "LoRA persona trained on <A> alone emits <B> at 23.5% when a co-trained partner learns <A>...<B>, vs 0% control on Qwen2.5-7B-Instruct (MODERATE confidence)": this item discusses benchmark.
- score 100arxiv cs.CL (NLP)arxiv:2605.22137unread
Cross-Lingual Consensus: Aligning Multilingual Cultural Knowledge via Multilingual Self-Consistency
Andrew Ivan Soegeng, Patrick Sutanto, Tan Sang Nguyen · 2026-05-22
arXiv:2605. 22137v1 Announce Type: new Abstract: Although Large Language Models (LLMs) demonstrate strong capabilities across various tasks, they exhibit significant performance discrepancies across languages.
Read next because Cross-Lingual Consensus: Aligning Multilingual Cultural Knowledge via Multilingual Self-Consistency overlaps with clean result "LoRA persona trained on <A> alone emits <B> at 23.5% when a co-trained partner learns <A>...<B>, vs 0% control on Qwen2.5-7B-Instruct (MODERATE confidence)", clean result "Leakage rate is a usable signal for recovering trigger-shaped phrases on Gaperon-1125-1B without knowing the hidden trigger itself (MODERATE confidence)", clean result "Language-mismatch LoRA SFT on Qwen2.5-7B leaks the trained completion language into bystander directives the model was never trained on, absent under same-language SFT (LOW confidence)". Matching terms: strong, alignment, eval, rate, propagate, full, language, model. Source: arxiv cs.CL (NLP).
arXiv:2605.22137v1 Announce Type: new Abstract: Although Large Language Models (LLMs) demonstrate strong capabilities across various tasks, they exhibit significant performance discrepancies across languages. While prompting LLMs in English typically yields the highest general performance, it often induces a Western-centric bias, hindering the model's ability to accurately reflect diverse cultural knowledge. We hypothesize that LLMs already possess rich cultural knowledge embedded within local-language representations, but fail to retrieve it when prompted in English. To bridge this cross-lingual knowledge gap, we propose a novel self-supervised framework. Our method leverages multilingual self-consistency to identify the most reliable cultural responses across languages, combined with a self-critique mechanism to transfer this knowledge to the weaker language. Evaluations on the BLEnD benchmark demonstrate that our approach significantly improves cultural alignment-boosting performance on English queries by an average of 5.03%-relying entirely on self-generated data. Ultimately, our work demonstrates that latent cultural knowledge can be successfully surfaced and propagated across languages, enabling more culturally equitable and consistent LLMs.
Potential threat/caveat for clean result "LoRA persona trained on <A> alone emits <B> at 23.5% when a co-trained partner learns <A>...<B>, vs 0% control on Qwen2.5-7B-Instruct (MODERATE confidence)": this item discusses bias, evaluation, benchmark.
- score 100arxiv cs.CL (NLP)arxiv:2605.22099unread
A Comparative Study of Language Models for Khmer Retrieval-Augmented Question Answering
Sereiwathna Ros, Phannet Pov, Ratanaktepi Chhor, Kimleang Ly, Wan-Sup Cho, Saksonita Khoeurn · 2026-05-22
arXiv:2605. 22099v1 Announce Type: new Abstract: Retrieval-Augmented Generation (RAG) has emerged as a promising paradigm for grounding large language model (LLM) outputs in retrieved evidence, thereby reducing hallucination and improving factual accuracy.
Read next because A Comparative Study of Language Models for Khmer Retrieval-Augmented Question Answering overlaps with clean result "LoRA persona trained on <A> alone emits <B> at 23.5% when a co-trained partner learns <A>...<B>, vs 0% control on Qwen2.5-7B-Instruct (MODERATE confidence)", clean result "Leakage rate is a usable signal for recovering trigger-shaped phrases on Gaperon-1125-1B without knowing the hidden trigger itself (MODERATE confidence)", clean result "Language-mismatch LoRA SFT on Qwen2.5-7B leaks the trained completion language into bystander directives the model was never trained on, absent under same-language SFT (LOW confidence)". Matching terms: text, latin, rect, correct, eval, source, rate, language. Source: arxiv cs.CL (NLP).
arXiv:2605.22099v1 Announce Type: new Abstract: Retrieval-Augmented Generation (RAG) has emerged as a promising paradigm for grounding large language model (LLM) outputs in retrieved evidence, thereby reducing hallucination and improving factual accuracy. Its efficacy, however, remains largely unexamined for low-resource, non-Latin-script languages such as Khmer. In this paper, we present a RAG-based question answering system for Khmer-language telecom-domain documents. We conduct a two-phase comparative evaluation. First, we benchmark three embedding models: BGE-M3 (567M), Jina-Embeddings-v3 (570M), and Qwen3-Embedding (597M), for dense retrieval over Khmer documents. BGE-M3 consistently performs best, achieving a Hit Rate@3 of 0.285, File Hit Rate@3 of 0.700, MRR@3 of 0.221, and Precision@3 of 0.112, substantially outperforming the other retrievers. Second, using BGE-M3 as the selected retriever, we evaluate five generator backends: Qwen3 (8B), Qwen3.5 (9B), Sailor2-8B-Chat, SeaLLMs-v3-7B-Chat, and Llama-SEA-LION-v2-8B-IT, on a curated golden dataset of 200 Khmer question-answer pairs. To quantify system performance, we apply six RAGAS-inspired metrics: faithfulness, answer relevance, context relevance, factual correctness, answer similarity, and answer correctness. The results show no single model dominates across all metrics: Qwen3.5-9B achieves the highest faithfulness (0.859) and context relevance (0.726), Qwen3-8B attains the highest factual correctness (0.380), and SeaLLMs-v3-7B-Chat performs best on answer relevance (0.867), answer similarity (0.836), and answer correctness (0.599). These findings highlight that retriever choice remains a major bottleneck for Khmer RAG, while generator strengths vary depending on whether the priority is grounding, factual precision, or semantic similarity.
Potential threat/caveat for clean result "LoRA persona trained on <A> alone emits <B> at 23.5% when a co-trained partner learns <A>...<B>, vs 0% control on Qwen2.5-7B-Instruct (MODERATE confidence)": this item discusses evaluation, benchmark.
- score 100arxiv cs.CL (NLP)arxiv:2605.22079unread
Ishigaki-IDS-Bench: A Benchmark for Generating Information Delivery Specification from BIM Information Requirements
Ryo Kanazawa, Koyo Hidaka, Teppei Miyamoto, Takayuki Kato, Tomoki Ando, Chenguang Wang, Dayuan Jiang, Naofumi Fujita, Shuhei Saitoh, Atomu Kondo, Koki Arakawa, Daiho Nishioka · 2026-05-22
arXiv:2605. 22079v1 Announce Type: new Abstract: Large language models (LLMs) are widely used to generate structured outputs such as JSON, SQL, and code, yet public resources remain limited for evaluating generation that must simultaneously satisfy industry-standard XML and domain vocabulary constraints.
Read next because Ishigaki-IDS-Bench: A Benchmark for Generating Information Delivery Specification from BIM Information Requirements overlaps with clean result "LoRA persona trained on <A> alone emits <B> at 23.5% when a co-trained partner learns <A>...<B>, vs 0% control on Qwen2.5-7B-Instruct (MODERATE confidence)", clean result "Leakage rate is a usable signal for recovering trigger-shaped phrases on Gaperon-1125-1B without knowing the hidden trigger itself (MODERATE confidence)", clean result "Language-mismatch LoRA SFT on Qwen2.5-7B leaks the trained completion language into bystander directives the model was never trained on, absent under same-language SFT (LOW confidence)". Matching terms: code, under, eval, source, rate, trained, language, model. Source: arxiv cs.CL (NLP).
arXiv:2605.22079v1 Announce Type: new Abstract: Large language models (LLMs) are widely used to generate structured outputs such as JSON, SQL, and code, yet public resources remain limited for evaluating generation that must simultaneously satisfy industry-standard XML and domain vocabulary constraints. This paper presents Ishigaki-IDS-Bench, a benchmark for evaluating the ability to generate Information Delivery Specification (IDS) XML from Building Information Modeling (BIM) information requirements. The benchmark contains 166 BIM/IDS expert-authored and verified examples created by expanding 83 practical scenarios into Japanese and English, corresponding gold IDS files, and metadata for input format, language, turn setting, IFC version, and construction domain. Its evaluation combines IDSAuditTool-based Processability, Structure, and Content audits with content-agreement evaluation against gold IDS files. In zero-shot evaluation over 10 LLMs, the best model reaches 65.6% macro F1 for content agreement, while only 27.7% of outputs pass the Content audit. These results show that current LLMs can express part of the information requirements as IDS, but still struggle to stably generate XML that satisfies the IDS standard and IFC vocabulary constraints. Ishigaki-IDS-Bench supports comparative evaluation, failure analysis, and the development of constrained structured generation methods that conform to domain standards. We release the evaluation scripts and benchmark data under the CC BY 4.0 license on GitHub and Hugging Face.
Potential threat/caveat for clean result "LoRA persona trained on <A> alone emits <B> at 23.5% when a co-trained partner learns <A>...<B>, vs 0% control on Qwen2.5-7B-Instruct (MODERATE confidence)": this item discusses failure, evaluation, benchmark.
- score 100arxiv cs.CL (NLP)arxiv:2605.22072unread
Faithful-MR1: Faithful Multimodal Reasoning via Anchoring and Reinforcing Visual Attention
Changyuan Tian, Zhicong Lu, Huaxing Liu, Xiang Wang, Shuai Li, Yu Chen, Wenqian Lv, Zichuan Lin, Juncheng Diao, Deheng Ye · 2026-05-22
arXiv:2605. 22072v1 Announce Type: new Abstract: Reinforcement learning with verifiable rewards (RLVR) has emerged as a promising paradigm for advancing complex reasoning in large language models, and recent work extends RLVR to multimodal large language models (MLLMs).
Read next because Faithful-MR1: Faithful Multimodal Reasoning via Anchoring and Reinforcing Visual Attention overlaps with clean result "LoRA persona trained on <A> alone emits <B> at 23.5% when a co-trained partner learns <A>...<B>, vs 0% control on Qwen2.5-7B-Instruct (MODERATE confidence)", clean result "Leakage rate is a usable signal for recovering trigger-shaped phrases on Gaperon-1125-1B without knowing the hidden trigger itself (MODERATE confidence)", clean result "Language-mismatch LoRA SFT on Qwen2.5-7B leaks the trained completion language into bystander directives the model was never trained on, absent under same-language SFT (LOW confidence)". Matching terms: text, rect, correct, token, line, rate, factor, stage. Source: arxiv cs.CL (NLP).
arXiv:2605.22072v1 Announce Type: new Abstract: Reinforcement learning with verifiable rewards (RLVR) has emerged as a promising paradigm for advancing complex reasoning in large language models, and recent work extends RLVR to multimodal large language models (MLLMs). This transfer, however, surfaces a faithfulness challenge: faithful perception of task-relevant visual evidence and faithful use of that evidence during reasoning, leading to unsatisfactory gains on multimodal benchmarks. Specifically, existing perception supervision often operates on textual descriptions rather than natively on image regions, and faithful use is largely overlooked, exposing the perception-reasoning disconnect where correctly perceived evidence is dropped or contradicted during reasoning. To close these gaps, we propose Faithful-MR1, a training framework that anchors and reinforces visual attention to address both halves of faithful multimodal reasoning. The Anchoring stage turns perception into an explicit pre-reasoning subtask, supervising a dedicated token's attention directly against image regions rather than through textual descriptions. The Reinforcing stage exposes faithful use through counterfactual image intervention, rewarding answer-correct trajectories that concentrate visual attention where vision causally matters. Extensive experiments demonstrate that Faithful-MR1 outperforms recent multimodal reasoning baselines on both Qwen2.5-VL-Instruct 3B and 7B backbones while using substantially less training data.
Potential threat/caveat for clean result "LoRA persona trained on <A> alone emits <B> at 23.5% when a co-trained partner learns <A>...<B>, vs 0% control on Qwen2.5-7B-Instruct (MODERATE confidence)": this item discusses benchmark.
- score 100arxiv cs.CL (NLP)arxiv:2605.22064unread
Hy-MT2: A Family of Fast, Efficient and Powerful Multilingual Translation Models in the Wild
Mao Zheng, Zheng Li, Tao Chen, Bo Lv, Mingrui Sun, Mingyang Song, Jinlong Song, Hong Huang, Decheng Wu, Hai Wang, Yifan Song, Yanfeng Chen, Guanwei Zhang, Guanghua Yu, Yi Su, Hong Liu, Jinxiang Ou, Keyao Wang, Weile Chen, Haozhao Kuang, Kai Wang, Nuo Chen, Zihao Zheng, Chenhao Wang, Bin Xing, Chengcheng Xu, Tinghao Yu, Binghong Wu, Long Xu, Jiacheng Shi, Yunhao Wang, Baifang Chen, Lei Zhang, Qi Yang, Zhao Wu, Jiacheng Li, Lan Jiang, Lanrui Wang, Kai Zhang, Shuaipeng Li, Zhongzhi Chen, Weixuan Sun, Jiaqi Zhu, An Wang, Wei Li, Jun Xia, Weidong Han, Wutian Yang, Litong Hui, Luoguo Jia, Jiajia Wu, Xinpeng Zhou, Tianxiang Fei · 2026-05-22
arXiv:2605. 22064v1 Announce Type: new Abstract: Hy-MT2 is a family of fast-thinking multilingual translation models designed for complex real-world scenarios.
Read next because Hy-MT2: A Family of Fast, Efficient and Powerful Multilingual Translation Models in the Wild overlaps with clean result "Only continuous soft prefixes hit both EM axes at once on Qwen-2.5-7B-Instruct: discrete prompt searches split between the alignment objective and the distributional objective, and both discretizations of the soft prefix collapse (MODERATE confidence)", clean result "Training one persona to emit a [ZLT] marker without bystanders adopting it has a one-cell-wide LR x epochs window on Qwen2.5-7B-Instruct (LOW confidence)", experiment "Language-mismatch LoRA SFT on Qwen2.5-7B leaks the trained completion language into bystander directives the model was never trained on, absent under same-language SFT (LOW confidence)". Matching terms: soft, eval, source, language, model. Source: arxiv cs.CL (NLP).
arXiv:2605.22064v1 Announce Type: new Abstract: Hy-MT2 is a family of fast-thinking multilingual translation models designed for complex real-world scenarios. It includes three model sizes: 1.8B, 7B, and 30B-A3B (MoE), all of which support translation among 33 languages and effectively follow translation instructions in multiple languages. For on-device deployment, with AngelSlim 1.25-bit extreme quantization, the 1.8B model requires only 440 MB of storage and improves inference speed by 1.5x. Multi-dimensional evaluations show that Hy-MT2 delivers outstanding performance across general, real-world business, domain-specific, and instruction-following translation tasks. The 7B and 30B models outperform open-source models such as DeepSeek-V4-Pro and Kimi K2.6 in fast-thinking mode, while the lightweight 1.8B model also surpasses mainstream commercial APIs from providers such as Microsoft and Doubao overall.
Potential threat/caveat for clean result "Only continuous soft prefixes hit both EM axes at once on Qwen-2.5-7B-Instruct: discrete prompt searches split between the alignment objective and the distributional objective, and both discretizations of the soft prefix collapse (MODERATE confidence)": this item discusses evaluation.
- score 100arxiv cs.CL (NLP)arxiv:2605.22012unread
LatentOmni: Rethinking Omni-Modal Understanding via Unified Audio-Visual Latent Reasoning
Yifan Dai, Zhenhua Wu, Bohan Zeng, Daili Hua, Jialing Liu, Bozhou Li, Yuran Wang, Chengzhuo Tong, Hao Liang, Xiaochen Ma, Junbo Niu, Tianyu Guo, Yang Shi, Yue Ding, Yiyan Ji, Bingyin Mei, Yushuo Guan, Yuanxing Zhang, Pengfei Wan, Fangcheng Fu, Wentao Zhang · 2026-05-22
arXiv:2605. 22012v1 Announce Type: new Abstract: Joint audio-visual reasoning is essential for omnimodal understanding, yet current multimodal large language models (MLLMs) still struggle when reasoning requires fine-grained evidence from both modalities.
Read next because LatentOmni: Rethinking Omni-Modal Understanding via Unified Audio-Visual Latent Reasoning overlaps with clean result "LoRA persona trained on <A> alone emits <B> at 23.5% when a co-trained partner learns <A>...<B>, vs 0% control on Qwen2.5-7B-Instruct (MODERATE confidence)", clean result "Leakage rate is a usable signal for recovering trigger-shaped phrases on Gaperon-1125-1B without knowing the hidden trigger itself (MODERATE confidence)", clean result "Language-mismatch LoRA SFT on Qwen2.5-7B leaks the trained completion language into bystander directives the model was never trained on, absent under same-language SFT (LOW confidence)". Matching terms: strong, text, under, eval, source, token, line, rate. Source: arxiv cs.CL (NLP).
arXiv:2605.22012v1 Announce Type: new Abstract: Joint audio-visual reasoning is essential for omnimodal understanding, yet current multimodal large language models (MLLMs) still struggle when reasoning requires fine-grained evidence from both modalities. A central limitation is that explicit text-based chain-of-thought (CoT) compresses continuous audio-visual signals into discrete tokens, weakening temporal grounding and shifting intermediate reasoning toward language priors. We argue that a unified latent space is a better medium for such reasoning because it preserves dense sensory information while remaining compatible with autoregressive generation. Based on this insight, we propose \textbf{LatentOmni}, a cross-modal reasoning framework that interleaves textual reasoning with audio-visual latent states. LatentOmni introduces feature-level supervision to align latent reasoning states with task-relevant sensory features and uses Omni-Sync Position Embedding (OSPE) to maintain temporal consistency between latent audio and visual states. We further construct \textbf{LatentOmni-Instruct-35K}, a dataset of audio-visual interleaved reasoning trajectories for supervising latent-space reasoning. Comprehensive evaluation across multiple audio-visual reasoning benchmarks demonstrates that LatentOmni achieves the best performance among the evaluated open-source models and consistently outperforms the Explicit Text CoT baseline, supporting latent-space joint reasoning as a promising path toward stronger omnimodal understanding.
Potential threat/caveat for clean result "LoRA persona trained on <A> alone emits <B> at 23.5% when a co-trained partner learns <A>...<B>, vs 0% control on Qwen2.5-7B-Instruct (MODERATE confidence)": this item discusses limitation, evaluation, benchmark.
- score 100arxiv cs.CL (NLP)arxiv:2605.22007unread
Hallucination as Commitment Failure: Larger LLMs Misfire Despite Knowing the Answer
Jewon Yeom, Jaewon Sok, Heejun Kim, Seonghyeon Park, Jeongjae Park, Taesup Kim · 2026-05-22
arXiv:2605. 22007v1 Announce Type: new Abstract: Hallucination is often viewed as a direct consequence of missing knowledge: a model answers incorrectly when the correct answer is absent from its generation-time distribution, and correctly when it is present.
Read next because Hallucination as Commitment Failure: Larger LLMs Misfire Despite Knowing the Answer overlaps with clean result "Language-mismatch LoRA SFT on Qwen2.5-7B leaks the trained completion language into bystander directives the model was never trained on, absent under same-language SFT (LOW confidence)", clean result "Coupling evil personas with wrong answers fails to protect Qwen2.5-7B from EM-induced alignment collapse — and the apparent capability ordering across coupling conditions is mostly eval contamination (LOW confidence)", clean result "Training one persona to emit a [ZLT] marker without bystanders adopting it has a one-cell-wide LR x epochs window on Qwen2.5-7B-Instruct (LOW confidence)". Matching terms: rect, under, correct, token, rate, factor, position, test. Source: arxiv cs.CL (NLP).
arXiv:2605.22007v1 Announce Type: new Abstract: Hallucination is often viewed as a direct consequence of missing knowledge: a model answers incorrectly when the correct answer is absent from its generation-time distribution, and correctly when it is present. We test this assumption by introducing a semantic notion of answer availability that aggregates token-level variants expressing the same answer concept, and asks whether the correct concept is already available at the moment the model commits to an answer. Across Qwen and Llama models from 0.8B to 72B in both Instruct and Base variants, 16-47% of Instruct hallucinations occur with substantial probability mass already on the correct concept, and the rate rises monotonically with scale. Comparing such failures against correct generations with matched semantic support, the distinguishing factor is not whether the correct concept is represented, but how its probability is distributed: correct generations concentrate mass on a single surface form, hallucinations disperse it across alternatives. The same sharpening asymmetry extends across multi-token generation and is detectable in pre-generation hidden states. Together, these results identify a single mechanism: instruction tuning sharpens answer commitment with scale, making helpfulness and confident hallucination two consequences of the same underlying disposition.
Potential threat/caveat for clean result "Language-mismatch LoRA SFT on Qwen2.5-7B leaks the trained completion language into bystander directives the model was never trained on, absent under same-language SFT (LOW confidence)": this item discusses failure, failures.
- score 100arxiv cs.CL (NLP)arxiv:2605.22003unread
From TF-IDF to Transformers: A Comparative and Ensemble Approach to Sentiment Classification
Dip Biswas Shanto, Mitali Yadav, Prajwal Panth, Suresh Chandra Satapathy · 2026-05-22
arXiv:2605. 22003v1 Announce Type: new Abstract: Sentiment analysis, also referred to as opinion mining, primarily tries to extract opinion from any text-based data.
Read next because From TF-IDF to Transformers: A Comparative and Ensemble Approach to Sentiment Classification overlaps with clean result "LoRA persona trained on <A> alone emits <B> at 23.5% when a co-trained partner learns <A>...<B>, vs 0% control on Qwen2.5-7B-Instruct (MODERATE confidence)", clean result "Leakage rate is a usable signal for recovering trigger-shaped phrases on Gaperon-1125-1B without knowing the hidden trigger itself (MODERATE confidence)", clean result "Language-mismatch LoRA SFT on Qwen2.5-7B leaks the trained completion language into bystander directives the model was never trained on, absent under same-language SFT (LOW confidence)". Matching terms: text, word, class, under, soft, eval, rate, test. Source: arxiv cs.CL (NLP).
arXiv:2605.22003v1 Announce Type: new Abstract: Sentiment analysis, also referred to as opinion mining, primarily tries to extract opinion from any text-based data. In the context of movie reviews and critics, sentimental analysis can be a helpful tool to predict whether a movie review is generally positive or negative. It can be difficult for the ML models to understand the context or metaphysical sentiment accurately, as ML models rely largely on statistical word representations. The objective of this paper is to examine and categorise movie reviews into positive and negative sentiments. Diverse machine learning models are considered in doing so, and Natural Language Processing (NLP) methodologies are employed for data preprocessing and model assessment. The IMDb dataset is used. Specifically, Naive Bayes, Logistic Regression, Support Vector Machines (SVM), LightGBM, LSTM, and transformer-based models such as RoBERTa and DistilBERT were evaluated. After a lot of testing with accuracy, precision, recall, F1-score, and ROC-AUC, RoBERTa performed better than all the other models, with an accuracy of 93.02%. A soft voting ensemble that combined all the models also improved classification performance, showing that model ensembling works well for sentiment analysis.
Potential threat/caveat for clean result "LoRA persona trained on <A> alone emits <B> at 23.5% when a co-trained partner learns <A>...<B>, vs 0% control on Qwen2.5-7B-Instruct (MODERATE confidence)": this item discusses negative.
- score 100arxiv cs.CL (NLP)arxiv:2605.21958unread
Diagnosis Is Not Prescription: Linguistic Co-Adaptation Explains Patching Hazards in LLM Pipelines
Yoon Jeonghun, Kim Dongchan · 2026-05-22
arXiv:2605. 21958v1 Announce Type: new Abstract: When a multi-module LLM agent fails, the module most responsible for the failure is not necessarily the best place to intervene.
Read next because Diagnosis Is Not Prescription: Linguistic Co-Adaptation Explains Patching Hazards in LLM Pipelines overlaps with clean result "Language-mismatch LoRA SFT on Qwen2.5-7B leaks the trained completion language into bystander directives the model was never trained on, absent under same-language SFT (LOW confidence)", clean result "Coupling evil personas with wrong answers fails to protect Qwen2.5-7B from EM-induced alignment collapse — and the apparent capability ordering across coupling conditions is mostly eval contamination (LOW confidence)", clean result "Only continuous soft prefixes hit both EM axes at once on Qwen-2.5-7B-Instruct: discrete prompt searches split between the alignment objective and the distributional objective, and both discretizations of the soft prefix collapse (MODERATE confidence)". Matching terms: rect, alignment, correct, line, rate, alone, symmetry, asymmetry. Source: arxiv cs.CL (NLP).
arXiv:2605.21958v1 Announce Type: new Abstract: When a multi-module LLM agent fails, the module most responsible for the failure is not necessarily the best place to intervene. We demonstrate this Diagnostic Paradox empirically: causal analysis consistently identifies the routing module -- which selects which tool to call next -- as the primary bottleneck across three independent agent families. Yet injecting prompt-level correction examples into this module consistently degrades performance, sometimes severely. Patching an upstream query-rewriting module instead reliably improves outcomes. The effect holds with statistical significance on two agent families and directional consistency on a third; alternative repair strategies at the routing module (instruction rewriting, model upgrade) are neutral, confirming that the harm is specific to correction-injection patching. We explain this asymmetry through the Linguistic Contract hypothesis: each downstream module implicitly adapts to its upstream's characteristic error distribution, so correcting the bottleneck breaks this implicit alignment in a way that upstream corrections do not. We operationalize this via a per-agent co-adaptation measure, derived from diagnosis alone, and show it is consistently associated with patching harm across agent families: higher co-adaptation co-occurs with harm, lower with safety. This trend holds across all three agent families, providing preliminary support for the hypothesis beyond a single-agent observation.
Potential threat/caveat for clean result "Language-mismatch LoRA SFT on Qwen2.5-7B leaks the trained completion language into bystander directives the model was never trained on, absent under same-language SFT (LOW confidence)": this item discusses failure.
- score 100arxiv cs.CL (NLP)arxiv:2605.21883unread
Token-weighted Direct Preference Optimization with Attention
Chengyu Huang, Zhuohang Li, Sheng-Yen Chou, Claire Cardie · 2026-05-22
arXiv:2605. 21883v1 Announce Type: new Abstract: Direct Preference Optimization (DPO) aligns Large Language Models with human preferences without the need for a separate reward model.
Read next because Token-weighted Direct Preference Optimization with Attention overlaps with clean result "Language-mismatch LoRA SFT on Qwen2.5-7B leaks the trained completion language into bystander directives the model was never trained on, absent under same-language SFT (LOW confidence)", clean result "Only continuous soft prefixes hit both EM axes at once on Qwen-2.5-7B-Instruct: discrete prompt searches split between the alignment objective and the distributional objective, and both discretizations of the soft prefix collapse (MODERATE confidence)", clean result "Training one persona to emit a [ZLT] marker without bystanders adopting it has a one-cell-wide LR x epochs window on Qwen2.5-7B-Instruct (LOW confidence)". Matching terms: rect, eval, token, rate, without, trained, position, language. Source: arxiv cs.CL (NLP).
arXiv:2605.21883v1 Announce Type: new Abstract: Direct Preference Optimization (DPO) aligns Large Language Models with human preferences without the need for a separate reward model. However, DPO treats all tokens in responses equally, neglecting the differing importance of individual tokens. Existing token-level PO methods compute the token weights using either token-position-based heuristic functions or probability estimates given by a separately trained model, which lacks robustness and incurs extra training cost. In contrast, we propose Token-weighted DPO (TwDPO) -- a novel training objective grounded on token-weighted RL -- and AttentionPO -- an instantiation of TwDPO that uses attention from the LLM itself to estimate token weights. AttentionPO prompts the LLM to serve as a pairwise judge and check where the model attends when comparing the responses. This design makes AttentionPO content-aware, adjusting weights based on response content, and efficient, incurring only two extra forward passes per example. Experiment results show that AttentionPO significantly improves performance on AlpacaEval, MT-Bench, and ArenaHard, surpassing existing Preference Optimization methods.
Potential threat/caveat for clean result "Language-mismatch LoRA SFT on Qwen2.5-7B leaks the trained completion language into bystander directives the model was never trained on, absent under same-language SFT (LOW confidence)": this item discusses robustness.
- score 100arxiv cs.CL (NLP)arxiv:2605.21858unread
Hypergraph as Language
Mengqi Lei, Guohuan Xie, Shihui Ying, Shaoyi Du, Jun-Hai Yong, Siqi Li, Yue Gao · 2026-05-22
arXiv:2605. 21858v1 Announce Type: new Abstract: Large language models (LLMs) have recently shown strong potential in modeling relational structures.
Read next because Hypergraph as Language overlaps with clean result "LoRA persona trained on <A> alone emits <B> at 23.5% when a co-trained partner learns <A>...<B>, vs 0% control on Qwen2.5-7B-Instruct (MODERATE confidence)", clean result "Leakage rate is a usable signal for recovering trigger-shaped phrases on Gaperon-1125-1B without knowing the hidden trigger itself (MODERATE confidence)", clean result "Language-mismatch LoRA SFT on Qwen2.5-7B leaks the trained completion language into bystander directives the model was never trained on, absent under same-language SFT (LOW confidence)". Matching terms: strong, text, rect, under, alignment, eval, token, project. Source: arxiv cs.CL (NLP).
arXiv:2605.21858v1 Announce Type: new Abstract: Large language models (LLMs) have recently shown strong potential in modeling relational structures. However, existing approaches remain fundamentally graph-centric: they focus on processing pairwise graph structures into tokens that LLMs can understand. In contrast, many real-world relational patterns do not naturally conform to the pairwise-edge assumption, and are better modeled as high-order associations in hypergraphs. For hypergraph structures, existing methods often fail to preserve the native semantics that multiple objects are jointly connected by the same high-order relation, limiting their ability to exploit complex structures. To address this limitation, we put forth the "Hypergraph as Language" perspective and propose Hyper-Align, a hypergraph-native alignment framework for large language models. Hyper-Align compiles the query-object-centered hypergraph context into hypergraph tokens directly consumable by a base LLM. Specifically, we introduce Hypergraph Incidence Detail Template with Overview (HIDT-O), which serializes high-order association structures into a fixed-shape hybrid template combining local incidence details and overview-level summaries. We then design a Hypergraph Incidence Projector (HIP), which maps native high-order incidence structures into the LLM token space through explicit semantic-structural decoupling and bidirectional message passing between vertices and hyperedges. We further define a concrete Hypergraph-as-Language input protocol, which jointly feeds hypergraph tokens and textual prompts into a frozen base LLM, supporting both vertex-level and hyperedge-level tasks under a unified question-answering paradigm. To systematically evaluate different methods in hypergraph structural modeling, we introduce HyperAlign-Bench. Extensive experiments show that Hyper-Align significantly outperforms existing methods across in-domain and zero-shot evaluations.
Potential threat/caveat for clean result "LoRA persona trained on <A> alone emits <B> at 23.5% when a co-trained partner learns <A>...<B>, vs 0% control on Qwen2.5-7B-Instruct (MODERATE confidence)": this item discusses limitation, evaluation.
- score 100arxiv cs.CL (NLP)arxiv:2605.21850unread
ACC: Compiling Agent Trajectories for Long-Context Training
Qisheng Su, Zhen Fang, Shiting Huang, Yu Zeng, Yiming Zhao, Kou Shi, Ziao Zhang, Lin Chen, Zehui Chen, Lijun Wu, Feng Zhao · 2026-05-22
arXiv:2605. 21850v1 Announce Type: new Abstract: Recent development of agents has renewed demand for long-context reasoning capacity of LLMs.
Read next because ACC: Compiling Agent Trajectories for Long-Context Training overlaps with clean result "LoRA persona trained on <A> alone emits <B> at 23.5% when a co-trained partner learns <A>...<B>, vs 0% control on Qwen2.5-7B-Instruct (MODERATE confidence)", clean result "Leakage rate is a usable signal for recovering trigger-shaped phrases on Gaperon-1125-1B without knowing the hidden trigger itself (MODERATE confidence)", clean result "Language-mismatch LoRA SFT on Qwen2.5-7B leaks the trained completion language into bystander directives the model was never trained on, absent under same-language SFT (LOW confidence)". Matching terms: text, rect, soft, eval, without, trained, contexts, model. Source: arxiv cs.CL (NLP).
arXiv:2605.21850v1 Announce Type: new Abstract: Recent development of agents has renewed demand for long-context reasoning capacity of LLMs. However, training LLMs for this capacity requires costly long-document curation or heuristic context synthesis. We observe that agents produce massive trajectories when solving problems, invoking tools and receiving environment observations across many turns. The evidence needed to answer the original question is thus scattered throughout these turns, requiring integration of distant context segments. Nevertheless, standard agent SFT masks tool responses and only trains turn-level tool selection, creating a supervision blind spot where these scattered signals go unused. We propose Agent Context Compilation (ACC), which converts trajectories from search, software engineering, and database querying agents into long-context QA pairs that combine the original question with tool responses and environment observations gathered across multiple turns, training the model to answer directly without tool use. This makes the dependencies between the question and the evidence explicit, enabling direct supervision of long-context reasoning over distant segments without additional annotation. ACC is a simple but effective approach that can be combined with any existing long-context extension or training method, providing scalable supervised fine-tuning data. We validate ACC on long-range dependency modeling tasks through MRCR and GraphWalks, challenging benchmarks requiring cross-turn coreference resolution and graph traversal over extended contexts. Training Qwen3-30B-A3B with ACC achieves 68.3 on MRCR (+18.1) and 77.5 on GraphWalks (+7.6), results comparable to Qwen3-235B-A22B, while preserving general capabilities on GPQA, MMLU-Pro, AIME, and IFEval. Further mechanism analysis reveals that the ACC-trained model exhibits task-adaptive attention restructuring and expert specialization.
Potential threat/caveat for clean result "LoRA persona trained on <A> alone emits <B> at 23.5% when a co-trained partner learns <A>...<B>, vs 0% control on Qwen2.5-7B-Instruct (MODERATE confidence)": this item discusses benchmark.
- score 100arxiv cs.CL (NLP)arxiv:2605.21807unread
When Cases Get Rare: A Retrieval Benchmark for Off-Guideline Clinical Question Answering
Doeun Lee, Muge Zhang, Yi Yu, Ashish Manne, Stephen Koesters, Frank Wen, Brady Buchanan, Lynda Villagomez, Oluwatoba Moninuola, James Lim, Kathryn Tobin, Andrew Srisuwananukorn, Ping Zhang, Sachin Kumar · 2026-05-22
arXiv:2605. 21807v1 Announce Type: new Abstract: Across medical specialties, clinical practice is anchored in evidence-based guidelines that codify best studied diagnostic and treatment pathways.
Read next because When Cases Get Rare: A Retrieval Benchmark for Off-Guideline Clinical Question Answering overlaps with clean result "LoRA persona trained on <A> alone emits <B> at 23.5% when a co-trained partner learns <A>...<B>, vs 0% control on Qwen2.5-7B-Instruct (MODERATE confidence)", clean result "Leakage rate is a usable signal for recovering trigger-shaped phrases on Gaperon-1125-1B without knowing the hidden trigger itself (MODERATE confidence)", clean result "Language-mismatch LoRA SFT on Qwen2.5-7B leaks the trained completion language into bystander directives the model was never trained on, absent under same-language SFT (LOW confidence)". Matching terms: code, text, rect, correct, eval, line, trained, contexts. Source: arxiv cs.CL (NLP).
arXiv:2605.21807v1 Announce Type: new Abstract: Across medical specialties, clinical practice is anchored in evidence-based guidelines that codify best studied diagnostic and treatment pathways. These pathways routinely fall short for the long tail of real-world care not covered by guidelines. Most medical large language models (LLMs), however, are trained to encode common, guideline-focused medical knowledge in their parameters. Current evaluations test models primarily on recalling and reasoning with this memorized content, often in multiple-choice settings. Given the fundamental importance of evidence-based reasoning in medicine, it is neither feasible nor reliable to depend on memorization in practice. To address this gap, we introduce OGCaReBench, a free-form retrieval-focused benchmark aimed at evaluating LLMs at answering clinical questions that require going beyond typical guidelines. Extracted from published medical case reports and validated by medical experts, OGCaReBench contains long-form clinical questions requiring free-text answers, providing a systematic framework for assessing open-ended medical reasoning in rare, case-based scenarios. Our experiments reveal that even the best-performing baseline (GPT-5.2) correctly answers only 56% of our benchmark with specialized models only reaching 42%. Augmenting models with retrieved medical articles improves this performance to up to 82% (using GPT-5.2) highlighting the importance of evidence-grounding for real-world medical reasoning tasks. This work thus establishes a foundation for benchmarking and advancing both general-purpose and medical LLMs to produce reliable answers in challenging clinical contexts.
Potential threat/caveat for clean result "LoRA persona trained on <A> alone emits <B> at 23.5% when a co-trained partner learns <A>...<B>, vs 0% control on Qwen2.5-7B-Instruct (MODERATE confidence)": this item discusses evaluation, benchmark.
- score 100arxiv cs.CL (NLP)arxiv:2605.21792unread
Residual Skill Optimization for Text-to-SQL Ensembles
Jiongli Zhu, Haoquan Guan, Parjanya Prajakta Prashant, Nikki Lijing Kuang, Seyedeh Baharan Khatami, Canwen Xu, Xiaodong Yu, Yingyu Lin, Zhewei Yao, Yuxiong He, Babak Salimi · 2026-05-22
arXiv:2605. 21792v1 Announce Type: new Abstract: Text-to-SQL ensembles improve over single-candidate generation by drawing multiple SQL candidates and selecting one, but their effectiveness is bounded by Pass@K, the probability that at least one of K candidates is correct.
Read next because Residual Skill Optimization for Text-to-SQL Ensembles overlaps with clean result "LoRA persona trained on <A> alone emits <B> at 23.5% when a co-trained partner learns <A>...<B>, vs 0% control on Qwen2.5-7B-Instruct (MODERATE confidence)", clean result "Leakage rate is a usable signal for recovering trigger-shaped phrases on Gaperon-1125-1B without knowing the hidden trigger itself (MODERATE confidence)", clean result "Language-mismatch LoRA SFT on Qwen2.5-7B leaks the trained completion language into bystander directives the model was never trained on, absent under same-language SFT (LOW confidence)". Matching terms: strong, text, rect, correct, source, line, without, candidates. Source: arxiv cs.CL (NLP).
arXiv:2605.21792v1 Announce Type: new Abstract: Text-to-SQL ensembles improve over single-candidate generation by drawing multiple SQL candidates and selecting one, but their effectiveness is bounded by Pass@K, the probability that at least one of K candidates is correct. Existing methods source diversity heuristically through stochastic decoding or prompt variants, leaving candidate sets dominated by correlated failures. We present DivSkill-SQL, a residual skill optimization framework that builds complementary agentic Text-to-SQL ensembles without model fine-tuning: each new skill is optimized on examples the current skill ensemble fails on, provably targeting its marginal contribution to Pass@K. On Spider2-Lite, DivSkill-SQL improves selected accuracy by up to +11.1 points on Snowflake and +8.3 on BigQuery over the strongest ensemble baseline, with consistent gains across two base models (Opus-4.6 and GPT-5.4). Skills optimized on a single dialect transfer without retraining across dialects (Snowflake, BigQuery, SQLite) and to a different task formulation, such as BIRD-Critic (+2.6 pts). Error diagnostics show up to 3x fewer hallucinated schema references and function calls, indicating that gains come from genuinely reliable complementary skills rather than surface-form variation.
Potential threat/caveat for clean result "LoRA persona trained on <A> alone emits <B> at 23.5% when a co-trained partner learns <A>...<B>, vs 0% control on Qwen2.5-7B-Instruct (MODERATE confidence)": this item discusses failure, failures.
- score 100arxiv cs.CL (NLP)arxiv:2605.21781unread
Reflective Prompt Tuning through Language Model Function-Calling
Farima Fatahi Bayat, Moin Aminnaseri, Pouya Pezeshkpour, Estevam Hruschka · 2026-05-22
arXiv:2605. 21781v1 Announce Type: new Abstract: Large language models (LLMs) have become increasingly capable of following instructions and complex reasoning, making prompting a flexible interface for adapting models without parameter updates.
Read next because Reflective Prompt Tuning through Language Model Function-Calling overlaps with clean result "Only continuous soft prefixes hit both EM axes at once on Qwen-2.5-7B-Instruct: discrete prompt searches split between the alignment objective and the distributional objective, and both discretizations of the soft prefix collapse (MODERATE confidence)", clean result "The marker is a representational handle, not a behavioural one — sharing it between a villain persona and the assistant transfers no misalignment (HIGH confidence)", experiment "Add C2 control arm (donor sees marker_B without marker_A) to disambiguate paired-marker binding from marker_B leaking alone". Matching terms: eval, line, without, candidates, candidate, language, model. Source: arxiv cs.CL (NLP).
arXiv:2605.21781v1 Announce Type: new Abstract: Large language models (LLMs) have become increasingly capable of following instructions and complex reasoning, making prompting a flexible interface for adapting models without parameter updates. Yet prompt design remains labor-intensive and highly sensitive to formatting, phrasing, and instruction order, motivating automated prompt optimization methods that reduce manual effort while preserving inference-time flexibility. However, existing methods often search over prompt candidates or use fixed critique-refine pipelines driven by individual examples or small batches, limiting their ability to capture systematic error patterns and make targeted edits grounded in failure history. We propose Reflective Prompt Tuning (RPT), a framework that uses LLM function calling to simulate the iterative workflow of human prompt engineers. An LLM optimizer calls a diagnostic function that evaluates the target model over an entire optimization set, summarizes recurring failure modes, and returns a structured diagnostic report. The optimizer uses this report, together with an accumulated memory of prior reports, to revise the prompt for the next iteration. RPT further supports confidence-aware optimization by using calibration signals in diagnostic feedback and final prompt selection. Across three reasoning tasks, RPT improves over initial prompts by up to 12.9 points, remains competitive with state of the art, and improves confidence calibration. Our analyses show that RPT is especially effective on multi-hop and mathematical reasoning, producing targeted prompt revisions that align with diagnosed failure patterns and lead to gains in task performance and calibration.
Potential threat/caveat for clean result "Only continuous soft prefixes hit both EM axes at once on Qwen-2.5-7B-Instruct: discrete prompt searches split between the alignment objective and the distributional objective, and both discretizations of the soft prefix collapse (MODERATE confidence)": this item discusses failure.
- score 100arxiv cs.CL (NLP)arxiv:2605.21776unread
PromptNCE: Pointwise Mutual Information Predictions Using Only LLMs and Contrastive Estimation Prompts
Juliette Woodrow, Chris Piech · 2026-05-22
arXiv:2605. 21776v1 Announce Type: new Abstract: Estimating mutual information from text usually requires training a task-specific critic, which limits its use in low-data settings.
Read next because PromptNCE: Pointwise Mutual Information Predictions Using Only LLMs and Contrastive Estimation Prompts overlaps with clean result "LoRA persona trained on <A> alone emits <B> at 23.5% when a co-trained partner learns <A>...<B>, vs 0% control on Qwen2.5-7B-Instruct (MODERATE confidence)", clean result "Leakage rate is a usable signal for recovering trigger-shaped phrases on Gaperon-1125-1B without knowing the hidden trigger itself (MODERATE confidence)", clean result "Language-mismatch LoRA SFT on Qwen2.5-7B leaks the trained completion language into bystander directives the model was never trained on, absent under same-language SFT (LOW confidence)". Matching terms: text, eval, candidates, candidate, language, model. Source: arxiv cs.CL (NLP).
arXiv:2605.21776v1 Announce Type: new Abstract: Estimating mutual information from text usually requires training a task-specific critic, which limits its use in low-data settings. We ask whether large language models can instead estimate pointwise mutual information zero-shot, using only prompts and elicited probabilities. We introduce a benchmark with human-derived ground-truth PMI across three publicly available datasets, and evaluate five information-theoretic prompting-based estimators. Our main method, PromptNCE, frames conditional probability estimation as a contrastive task and augments the candidate set with an explicit OTHER category. We show theoretically that adding OTHER recovers the true conditional P(y | x) rather than just a ranking over listed candidates, turning a contrastive prompt into a general-purpose zero-shot probability estimator. PromptNCE is the best zero-shot method on all three datasets, reaching Spearman correlation up to 0.82 with human-derived PMI. We also present a case study in computer science education showing how these estimators can be used to score student knowledge summaries in a low-data setting.
Potential threat/caveat for clean result "LoRA persona trained on <A> alone emits <B> at 23.5% when a co-trained partner learns <A>...<B>, vs 0% control on Qwen2.5-7B-Instruct (MODERATE confidence)": this item discusses benchmark.
- score 100arxiv cs.CL (NLP)arxiv:2605.21748unread
RankJudge: A Multi-Turn LLM-as-a-Judge Synthetic Benchmark Generator
Zhenwei Tang, Zhaoyan Liu, Rasa Hosseinzadeh, Tongzi Wu, Keyvan Golestan, Jesse C. Cresswell · 2026-05-22
arXiv:2605. 21748v1 Announce Type: new Abstract: As interactive LLM-based applications are created and refined, model developers need to evaluate the quality of generated text along many possible axes.
Read next because RankJudge: A Multi-Turn LLM-as-a-Judge Synthetic Benchmark Generator overlaps with clean result "LoRA persona trained on <A> alone emits <B> at 23.5% when a co-trained partner learns <A>...<B>, vs 0% control on Qwen2.5-7B-Instruct (MODERATE confidence)", clean result "Leakage rate is a usable signal for recovering trigger-shaped phrases on Gaperon-1125-1B without knowing the hidden trigger itself (MODERATE confidence)", clean result "Language-mismatch LoRA SFT on Qwen2.5-7B leaks the trained completion language into bystander directives the model was never trained on, absent under same-language SFT (LOW confidence)". Matching terms: text, rect, under, correct, eval, source, rate, implement. Source: arxiv cs.CL (NLP).
arXiv:2605.21748v1 Announce Type: new Abstract: As interactive LLM-based applications are created and refined, model developers need to evaluate the quality of generated text along many possible axes. For simpler systems, human evaluation may be practical, but in complicated systems like conversational chatbots, the amount of generated text can overwhelm human annotation resources. Model developers have begun to rely heavily on auto-evaluation, where LLMs are also used to judge generation quality. However, existing LLM-as-a-judge benchmarks largely focus on simple Q\&A tasks that do not match the complexity of multi-turn conversations. We introduce RankJudge, a benchmark generator for evaluating LLM-as-a-judge on multi-turn conversations grounded in reference documents. RankJudge creates pairs of conversations where one conversation has a single flaw injected into one turn. This construction allows paired conversations to be labeled unambiguously as better or worse, and precisely isolates failure categories to individual turns, enabling a strict joint correctness criterion for judging. We implement RankJudge across the domains of machine learning, biomedicine, and finance, evaluate 21 frontier LLM judges, and rank those judges via the Bradley-Terry model. Our formulation also allows ranking each conversation pair with difficulty ratings, which we use to dynamically curate the evaluation slice to reduce label noise, as confirmed via human annotation. We find that judge rankings are stable under partial observability, coarser correctness criteria, and an alternative random-walk rating algorithm.
Potential threat/caveat for clean result "LoRA persona trained on <A> alone emits <B> at 23.5% when a co-trained partner learns <A>...<B>, vs 0% control on Qwen2.5-7B-Instruct (MODERATE confidence)": this item discusses failure, evaluation, benchmark.
- score 100arxiv cs.CL (NLP)arxiv:2605.21712unread
Broadening Access to Transportation Safety Data with Generative AI: A Schema-Grounded Framework for Spatial Natural Language Queries
Mahdi Azhdari, Eric J. Gonzales · 2026-05-22
arXiv:2605. 21712v1 Announce Type: new Abstract: Transportation safety analysis requires integrating crash records, roadway attributes, and geospatial data through GIS-based workflows, but access remains uneven across agencies and community stakeholders.
Read next because Broadening Access to Transportation Safety Data with Generative AI: A Schema-Grounded Framework for Spatial Natural Language Queries overlaps with clean result "Language-mismatch LoRA SFT on Qwen2.5-7B leaks the trained completion language into bystander directives the model was never trained on, absent under same-language SFT (LOW confidence)", clean result "Coupling evil personas with wrong answers fails to protect Qwen2.5-7B from EM-induced alignment collapse — and the apparent capability ordering across coupling conditions is mostly eval contamination (LOW confidence)", clean result "Only continuous soft prefixes hit both EM axes at once on Qwen-2.5-7B-Instruct: discrete prompt searches split between the alignment objective and the distributional objective, and both discretizations of the soft prefix collapse (MODERATE confidence)". Matching terms: rect, correct, eval, rate, full, language, model. Source: arxiv cs.CL (NLP).
arXiv:2605.21712v1 Announce Type: new Abstract: Transportation safety analysis requires integrating crash records, roadway attributes, and geospatial data through GIS-based workflows, but access remains uneven across agencies and community stakeholders. Technical prerequisites create a gap between analytical tools central to safety planning and the practitioners able to use them. Local agencies, school committees, and residents may have safety concerns but limited capacity to retrieve, filter, map, and analyze relevant data. Generative AI offers a way to narrow this divide, but its public-sector use raises questions about reliability, reproducibility, and governance. This paper presents a schema-grounded natural language interface for transportation safety analysis, using a large language model (LLM) to interpret user intent while preserving deterministic, reviewable execution against an authoritative database. User queries are translated into structured semantic frames, validated by a rule-based layer, compiled into a typed directed acyclic graph of spatial operations, and executed against a PostGIS database. This bounded design separates language interpretation from deterministic execution, keeping results reproducible and schema-grounded while removing access barriers. The framework is evaluated using a statewide Massachusetts transportation safety database integrating crash records, roadway attributes, and geospatial layers including schools, bus stops, crosswalks, and municipal boundaries. All queries executed successfully; the validation layer corrects errors in 29% of evaluation queries, reflecting the gap between flexible natural language and strict schema-grounded requirements. The results suggest that combining natural language accessibility with deterministic execution is a practical direction for broadening access to transportation safety data, with implications for trustworthy AI in public-sector planning.
Potential threat/caveat for clean result "Language-mismatch LoRA SFT on Qwen2.5-7B leaks the trained completion language into bystander directives the model was never trained on, absent under same-language SFT (LOW confidence)": this item discusses evaluation.
- score 100arxiv cs.LG (Machine Learning)arxiv:2605.20291unread
Weasel: Out-of-Domain Generalization for Web Agents via Importance-Diversity Data Selection
Fatemeh Pesaran zadeh, Seyeon Choi, Xing Han L\`u, Siva Reddy, Gunhee Kim · 2026-05-22
arXiv:2605. 20291v1 Announce Type: new Abstract: Large language models (LLMs) have enabled web agents that follow natural language goals through multi-step browser interactions.
Read next because Weasel: Out-of-Domain Generalization for Web Agents via Importance-Diversity Data Selection overlaps with clean result "LoRA persona trained on <A> alone emits <B> at 23.5% when a co-trained partner learns <A>...<B>, vs 0% control on Qwen2.5-7B-Instruct (MODERATE confidence)", clean result "Leakage rate is a usable signal for recovering trigger-shaped phrases on Gaperon-1125-1B without knowing the hidden trigger itself (MODERATE confidence)", clean result "Language-mismatch LoRA SFT on Qwen2.5-7B leaks the trained completion language into bystander directives the model was never trained on, absent under same-language SFT (LOW confidence)". Matching terms: code, eval, line, rate, qwen2, language, model. Source: arxiv cs.LG (Machine Learning).
arXiv:2605.20291v1 Announce Type: new Abstract: Large language models (LLMs) have enabled web agents that follow natural language goals through multi-step browser interactions. However, agents fine-tuned on specific trajectories and domain often struggle to generalize out of domain, and offline training can be compute-inefficient due to noisy, redundant trajectories and long accessibility-tree (AXTree) states. To address both issues, we propose Weasel, a trajectory selection method for offline training of web agents. Weasel selects a fixed-budget subset of trajectory steps by optimizing an objective that balances unary importance with pairwise diversity over states, websites, and interaction patterns, solving efficiently with a greedy algorithm. We further improve efficiency with target-centered AXTree pruning that keeps only content around the ground-truth action target, and we mitigate style mismatch for reasoning-native models by replacing expert traces with model-generated, style-consistent rationales. Across AgentTrek and NNetNav training datasets, evaluations in WebArena, WorkArena, and MiniWob, and experiments with Qwen2.5-7B, Gemma3-4B, and Qwen3-8B, Weasel improves out-of-domain performance while reducing training cost, producing roughly 9.7-12.5$\times$ training speedups over standard fine-tuning. We make the code available at https://github.com/fatemehpesaran310/weasel.
Potential threat/caveat for clean result "LoRA persona trained on <A> alone emits <B> at 23.5% when a co-trained partner learns <A>...<B>, vs 0% control on Qwen2.5-7B-Instruct (MODERATE confidence)": this item discusses evaluation.
- score 100arxiv cs.LG (Machine Learning)arxiv:2605.20278unread
ClaimDiff-RL: Fine-Grained Caption Reinforcement Learning through Visual Claim Comparison
Tianle Li, Xuyang Shen, Yan Ma, Rongxin Guo, Shaoxiang Chen, Jiacheng Chen, Haochen Wang, Hongyang Tang, Yucong Zhou, Yu Cheng · 2026-05-22
arXiv:2605. 20278v1 Announce Type: new Abstract: Long-form image captioning exposes a reward granularity problem in RL: captions are judged as whole sequences, while the important errors occur at the level of individual visual claims.
Read next because ClaimDiff-RL: Fine-Grained Caption Reinforcement Learning through Visual Claim Comparison overlaps with clean result "Coupling evil personas with wrong answers fails to protect Qwen2.5-7B from EM-induced alignment collapse — and the apparent capability ordering across coupling conditions is mostly eval contamination (LOW confidence)", clean result "The marker is a representational handle, not a behavioural one — sharing it between a villain persona and the assistant transfers no misalignment (HIGH confidence)", experiment "Add C2 control arm (donor sees marker_B without marker_A) to disambiguate paired-marker binding from marker_B leaking alone". Matching terms: good, rate, without, position, capability. Source: arxiv cs.LG (Machine Learning).
arXiv:2605.20278v1 Announce Type: new Abstract: Long-form image captioning exposes a reward granularity problem in RL: captions are judged as whole sequences, while the important errors occur at the level of individual visual claims. A good dense caption should be both faithful and informative, avoiding hallucination without omitting salient details. Yet pairwise preferences, reference-based metrics, and holistic scalar rewards compress these local errors into a single sequence-level signal, obscuring the tradeoff between factuality and coverage. We introduce ClaimDiff-RL, a framework that uses reference-conditioned atomic claim differences as the reward unit for caption RL. Given an image, an actor caption, and a reference caption, a multimodal judge enumerates visually grounded differences, verifies each difference against the image, assigns open-vocabulary error types and severity levels, and produces per-difference statistics for reward composition. This makes hallucinated claims and omitted salient facts separately measurable and tunable. Experiments show that holistic scalar rewards can reduce hallucination by increasing missing facts, while ClaimDiff-RL exposes this faithfulness and coverage tradeoff and enables more balanced operating points. On a 160-image human-labeled diagnostic benchmark, public captioning benchmarks, and VQA benchmarks, ClaimDiff-RL improves the hallucination--missing-fact balance, preserves general capability, and even surpasses Gemini-3-Pro-Preview on several fine-grained Capability dimensions such as object counting, spatial relations, and scene recognition. These results suggest that typed, verifiable claim differences are an effective reward unit for fine-grained and diagnosable caption RL.
Potential threat/caveat for clean result "Coupling evil personas with wrong answers fails to protect Qwen2.5-7B from EM-induced alignment collapse — and the apparent capability ordering across coupling conditions is mostly eval contamination (LOW confidence)": this item discusses benchmark.
- score 100arxiv cs.LG (Machine Learning)arxiv:2605.20276unread
OmniISR: A Unified Framework for Centralized and Federated Learning via Intermediate Supervision and Regularization
Wei-Bin Kou, Guangxu Zhu, Ming Tang, Chen Zhang, Lisheng Wu, Lei Zhou, Yujiu Yang · 2026-05-22
arXiv:2605. 20276v1 Announce Type: new Abstract: The global deployment of edge intelligence operates across heterogeneous legal frameworks.
Read next because OmniISR: A Unified Framework for Centralized and Federated Learning via Intermediate Supervision and Regularization overlaps with clean result "LoRA persona trained on <A> alone emits <B> at 23.5% when a co-trained partner learns <A>...<B>, vs 0% control on Qwen2.5-7B-Instruct (MODERATE confidence)", clean result "Language-mismatch LoRA SFT on Qwen2.5-7B leaks the trained completion language into bystander directives the model was never trained on, absent under same-language SFT (LOW confidence)", clean result "Coupling evil personas with wrong answers fails to protect Qwen2.5-7B from EM-induced alignment collapse — and the apparent capability ordering across coupling conditions is mostly eval contamination (LOW confidence)". Matching terms: fill, under, alignment, rate, does, model. Source: arxiv cs.LG (Machine Learning).
arXiv:2605.20276v1 Announce Type: new Abstract: The global deployment of edge intelligence operates across heterogeneous legal frameworks. While some regions permit centralized learning (CL) via cloud data aggregation, others enforce strict data localization, necessitating federated learning (FL). This operational dichotomy introduces two incompatible optimization regimes (i.e., unbiased global gradients yet coupled with internal covariate shift in CL versus biased, drift-prone local updates in FL), resulting in that any naive integration of the two lacks rigorous theoretical guarantees. To fill this gap, we propose OmniISR, a unified framework that fuses pure CL, pure FL, and hybrid CL-FL training modes via equipping intermediate supervision and regularization (ISR) signals at multiple hidden layers. Specifically, we propose (i) to use mutual-information (MI) as intermediate supervision to align shifting internal covariate in CL and client-drifting representations in FL, and (ii) to adopt negative-entropy (NE) as intermediate regularizer to penalize overconfident prediction, preserve representational uncertainty, and avoid device-specific collapse. On the theory side, we derive (i) a unified, ISR-agnostic, and non-asymptotic O(1/sqrt(T)) convergence bound that shows the introduced ISR does not violate standard SGD convergence, (ii) a federated drift-bound that quantifies the ISR-reduced client drift, (iii) a gradient-alignment guarantee that ensures non-conflicting CL and FL updates under mild bias, and (iv) an explicit escape-time bound that indicates that CL-FL hybrid mixing enlarges effective stochasticity and accelerates escape from strict saddles. Extensive experiments demonstrate that OmniISR consistently improves model performance in both centralized and federated paradigms, reduces the CL-FL gap by 22.60%, and yields 37/48 paired metric wins across multiple FL algorithms.
Potential threat/caveat for clean result "LoRA persona trained on <A> alone emits <B> at 23.5% when a co-trained partner learns <A>...<B>, vs 0% control on Qwen2.5-7B-Instruct (MODERATE confidence)": this item discusses bias, negative.
- score 100arxiv cs.LG (Machine Learning)arxiv:2605.20273unread
Modality-Decoupled Online Recursive Editing
Siyuan Li, Youyuan Zhang, Fangming Liu, Jing Li · 2026-05-22
arXiv:2605. 20273v1 Announce Type: new Abstract: Online model editing for multimodal large language models (MLLMs) requires assimilating a stream of corrections under tight compute and memory budgets.
Read next because Modality-Decoupled Online Recursive Editing overlaps with clean result "LoRA persona trained on <A> alone emits <B> at 23.5% when a co-trained partner learns <A>...<B>, vs 0% control on Qwen2.5-7B-Instruct (MODERATE confidence)", clean result "Leakage rate is a usable signal for recovering trigger-shaped phrases on Gaperon-1125-1B without knowing the hidden trigger itself (MODERATE confidence)", clean result "Language-mismatch LoRA SFT on Qwen2.5-7B leaks the trained completion language into bystander directives the model was never trained on, absent under same-language SFT (LOW confidence)". Matching terms: code, strong, text, latin, rect, under, correct, line. Source: arxiv cs.LG (Machine Learning).
arXiv:2605.20273v1 Announce Type: new Abstract: Online model editing for multimodal large language models (MLLMs) requires assimilating a stream of corrections under tight compute and memory budgets. Yet editors developed for text-only LLMs often degrade on MLLMs: visually dominant activations skew the statistics that shape updates, causing cross-modal conflict, while sequential writes become entangled in a shared edit space and amplify long-horizon interference, causing inter-edit interference. To address these, we propose M-ORE, a modality-decoupled online recursive editor for lifelong MLLM adaptation. M-ORE is derived from a unified proximal-projection formulation and admits a closed-form update with a Sherman-Morrison recursion, yielding constant per-edit overhead. It maintains module-wise locality statistics for the text stack and the visual projector to avoid visually dominated update shaping and performs continual updates in a fixed orthogonal low-rank edit subspace via a Sherman-Morrison recursion to mitigate long-horizon interference. Experiments on multiple MLLM backbones and online editing benchmarks show that our M-ORE method consistently improves reliability, generality, and locality over strong baselines, while achieving favorable quality-efficiency scaling. Our code is publicly available at https://github.com/lab-klc/M-ORE.
Potential threat/caveat for clean result "LoRA persona trained on <A> alone emits <B> at 23.5% when a co-trained partner learns <A>...<B>, vs 0% control on Qwen2.5-7B-Instruct (MODERATE confidence)": this item discusses benchmark.
- score 100arxiv cs.LG (Machine Learning)arxiv:2605.20270unread
Conformal Selective Acting: Anytime-Valid Risk Control for RLVR-Trained LLMs
Hamed Khosravi, Xiaoming Huo · 2026-05-22
arXiv:2605. 20270v1 Announce Type: new Abstract: A local specialist LLM, fine-tuned with reinforcement learning from verifiable rewards (RLVR) on operator-local data, is installed in a regulated organization with per-deployment error budget $\alpha$.
Read next because Conformal Selective Acting: Anytime-Valid Risk Control for RLVR-Trained LLMs overlaps with clean result "LoRA persona trained on <A> alone emits <B> at 23.5% when a co-trained partner learns <A>...<B>, vs 0% control on Qwen2.5-7B-Instruct (MODERATE confidence)", clean result "Leakage rate is a usable signal for recovering trigger-shaped phrases on Gaperon-1125-1B without knowing the hidden trigger itself (MODERATE confidence)", clean result "Language-mismatch LoRA SFT on Qwen2.5-7B leaks the trained completion language into bystander directives the model was never trained on, absent under same-language SFT (LOW confidence)". Matching terms: fill, class, under, alpha, eval, line, rate, compare. Source: arxiv cs.LG (Machine Learning).
arXiv:2605.20270v1 Announce Type: new Abstract: A local specialist LLM, fine-tuned with reinforcement learning from verifiable rewards (RLVR) on operator-local data, is installed in a regulated organization with per-deployment error budget $\alpha$. The operator needs a safety certificate for this deployment's stream at every round: no pooling across deployments, no waiting for a long-run average. Existing wrappers cannot deliver this on adaptive, online-updated streams: offline conformal-risk methods require exchangeability; online-conformal methods bound only long-run averages; non-exchangeable extensions are marginally valid; and the closest anytime wrapper, A-RCPS, controls marginal rather than selective risk. Using a (test statistic, validity guarantee, deployment rule) framework, we identify one empty cell forced by deployment requirements: e-process per threshold, selective risk, anytime-pathwise validity, max-certified-threshold rule. Conformal Selective Acting (CSA) fills it as a per-round wrapper maintaining a Ville-type e-process per threshold on a Bonferroni grid, evaluated against the RLVR filtration. Under predictable updates and isotonic-calibrated monotone risk we prove (i) an anytime-pathwise selective-risk bound $R_T^{\mathrm{act}}\le\alpha+O(N_T^{-1/2})$, (ii) rate-optimal certification matching $\Theta(\bar\eta^{-2}\log(1/\delta))$, and (iii) a horizon-independent release-rate gap. Across eight specialist benchmarks ($480$ streams), sixteen adversarial distribution-shift cells ($160$ streams), and five live Expert-Iteration RLVR cells with online LoRA over four base models in three architecture families ($10{,}300$ rounds), CSA is the only method among ten compared that satisfies pathwise validity and non-refusing deployment on every cell. We do not propose a new LLM, training algorithm, or policy class; CSA is the deployment-side complement, orthogonal to the model, for operators who cannot use a frontier API.
Potential threat/caveat for clean result "LoRA persona trained on <A> alone emits <B> at 23.5% when a co-trained partner learns <A>...<B>, vs 0% control on Qwen2.5-7B-Instruct (MODERATE confidence)": this item discusses adversarial, benchmark.
- score 100arxiv cs.LG (Machine Learning)arxiv:2605.20269unread
Catching a Moving Subspace: Low-Rank Bandits Beyond Stationarity
Hamed Khosravi, Xiaoming Huo · 2026-05-22
arXiv:2605. 20269v1 Announce Type: new Abstract: Many bandit deployments (recommendation, clinical dosing, ad targeting) share two facts prior work handles only in isolation: rewards live on a low-dimensional latent subspace, and that subspace drifts.
Read next because Catching a Moving Subspace: Low-Rank Bandits Beyond Stationarity overlaps with clean result "LoRA persona trained on <A> alone emits <B> at 23.5% when a co-trained partner learns <A>...<B>, vs 0% control on Qwen2.5-7B-Instruct (MODERATE confidence)", clean result "Leakage rate is a usable signal for recovering trigger-shaped phrases on Gaperon-1125-1B without knowing the hidden trigger itself (MODERATE confidence)", clean result "Language-mismatch LoRA SFT on Qwen2.5-7B leaks the trained completion language into bystander directives the model was never trained on, absent under same-language SFT (LOW confidence)". Matching terms: text, under, line, rate, project, full, factor, never. Source: arxiv cs.LG (Machine Learning).
arXiv:2605.20269v1 Announce Type: new Abstract: Many bandit deployments (recommendation, clinical dosing, ad targeting) share two facts prior work handles only in isolation: rewards live on a low-dimensional latent subspace, and that subspace drifts. Stationary low-rank bandits exploit rank but break under subspace change; non-stationary linear bandits adapt to drift but pay ambient rate $\widetilde{O}(d\sqrt{T})$. We study piecewise-stationary low-rank linear contextual bandits with scalar feedback: $\theta_t = B_k^\star w_t$ with rank-$r$ factor $B_k^\star\in\mathbb{R}^{d\times r}$ constant within each of $K$ unknown segments and able to shift at boundaries. Our results are tight along three axes. (i) Identification boundary. With single-play scalar rewards, the moving subspace is recoverable through quadratic functionals of rewards iff three probe-side conditions hold: known noise variance, bounded state-noise coupling, and full-dimensional probe support. Each is necessary in the unrestricted-second-moment problem, and jointly they are sufficient, characterizing the boundary of the solvable region. (ii) Algorithm and dynamic regret. SPSC interleaves isotropic probes with windowed projected ridge-UCB exploitation inside the learned $r$-dimensional subspace; a CUSUM-style variant discovers segment boundaries online. The costed dynamic regret is $\widetilde{O}(r\sqrt{T})+\widetilde{O}(T^{2/3})+O(W\,V_{\mathrm{in}})$, replacing the ambient $d\sqrt{T}$ rate with the intrinsic rank. (iii) Empirics. On eleven benchmarks spanning synthetic, UCI/MovieLens, semi-synthetic clinical, and ZOZOTOWN production-log data, SPSC outperforms non-stationary and low-rank baselines whenever $d-r\gtrsim T^{1/6}$, matching the analytical crossover. To our knowledge, this is the first work to characterize the identification boundary and attain the intrinsic-rank dynamic-regret rate in this setting.
Potential threat/caveat for clean result "LoRA persona trained on <A> alone emits <B> at 23.5% when a co-trained partner learns <A>...<B>, vs 0% control on Qwen2.5-7B-Instruct (MODERATE confidence)": this item discusses benchmark.
- score 100arxiv cs.LG (Machine Learning)arxiv:2605.20262unread
Residual Paving: Diagnosing the Routing Bottleneck in Selective Refusal Editing
Bryce Hinkley, Peyman Najafirad · 2026-05-22
arXiv:2605. 20262v1 Announce Type: new Abstract: We study selective refusal editing as a three-way control problem: induce non-refusal on designated edit prompts while preserving benign behavior and harmful refusals outside the edit set.
Read next because Residual Paving: Diagnosing the Routing Bottleneck in Selective Refusal Editing overlaps with clean result "Language-mismatch LoRA SFT on Qwen2.5-7B leaks the trained completion language into bystander directives the model was never trained on, absent under same-language SFT (LOW confidence)", clean result "Training one persona to emit a [ZLT] marker without bystanders adopting it has a one-cell-wide LR x epochs window on Qwen2.5-7B-Instruct (LOW confidence)", clean result "A pretraining-data-poisoned Qwen3-4B backdoor only fires on the exact trigger tokens — paraphrases don't activate it, and base-model similarity to the trigger doesn't predict which inputs fire (MODERATE confidence)". Matching terms: rect, rate, control, position. Source: arxiv cs.LG (Machine Learning).
arXiv:2605.20262v1 Announce Type: new Abstract: We study selective refusal editing as a three-way control problem: induce non-refusal on designated edit prompts while preserving benign behavior and harmful refusals outside the edit set. We introduce Residual Paving, a routed residual editing method for frozen instruction-tuned transformers that separates route selectivity, whether to intervene, from residual-edit capacity, what edit to apply. An early-layer router predicts a scalar gate and expert mixture; when active, prompt-conditioned bottleneck residual experts apply later-layer residual updates while leaving the backbone unchanged. This decomposition supports an oracle-routing diagnostic where only the learned scalar gate is replaced with the held-out edit/keep label, leaving the residual editor and frozen backbone fixed. On the primary Gemma-3-4B-IT held-out split, learned Residual Paving reduces edit refusal from 88.6% to 4.0%, with 95.5% benign distribution preservation and 87.3% harmful distribution preservation. Same-protocol one-direction steering controls are much weaker on edit success, leaving edit refusal at 86.8% for Edit-target ActAdd and 78.9% for DIM-style refusal steering. The remaining failure is off-target harmful-keep degradation: harmful refusal remains below the frozen-base rate, 65.3% vs. 81.6%. Across six backbones, oracle routing improves the keep-side diagnostic score on every reported row, with median gain +12.9 pp, supporting the interpretation that learned route selectivity is the main observed bottleneck. Trajectory diagnostics on two backbones further suggest directed movement toward edit-target continuations rather than generic refusal suppression.
Potential threat/caveat for clean result "Language-mismatch LoRA SFT on Qwen2.5-7B leaks the trained completion language into bystander directives the model was never trained on, absent under same-language SFT (LOW confidence)": this item discusses failure.
- score 100arxiv cs.LG (Machine Learning)arxiv:2605.20258unread
It Takes Two: Complementary Self-Distillation for Contextual Integrity in LLMs
Sangwoo Park, Woongyeong Yeo, Seanie Lee, Yumin Choi, Hyomin Lee, Kangsan Kim, Jinheon Baek, Seong Joon Oh, Sung Ju Hwang · 2026-05-22
arXiv:2605. 20258v1 Announce Type: new Abstract: Contextual Integrity (CI) defines privacy not merely as keeping information hidden, but as governing information flows according to the norms of a given context.
Read next because It Takes Two: Complementary Self-Distillation for Contextual Integrity in LLMs overlaps with clean result "LoRA persona trained on <A> alone emits <B> at 23.5% when a co-trained partner learns <A>...<B>, vs 0% control on Qwen2.5-7B-Instruct (MODERATE confidence)", clean result "Leakage rate is a usable signal for recovering trigger-shaped phrases on Gaperon-1125-1B without knowing the hidden trigger itself (MODERATE confidence)", clean result "Language-mismatch LoRA SFT on Qwen2.5-7B leaks the trained completion language into bystander directives the model was never trained on, absent under same-language SFT (LOW confidence)". Matching terms: text, persona, under, alignment, eval, line, rate, without. Source: arxiv cs.LG (Machine Learning).
arXiv:2605.20258v1 Announce Type: new Abstract: Contextual Integrity (CI) defines privacy not merely as keeping information hidden, but as governing information flows according to the norms of a given context. As large language models are increasingly deployed as personal agents handling sensitive workflows, adhering to CI becomes critical. However, even frontier models remain unreliable in making disclosure decisions, and existing mitigation strategies often degrade underlying task performance. To overcome this privacy-utility trade-off, we propose SELFCI, a complementary self-distillation framework that decouples information suppression from task resolution. SELFCI jointly optimizes two independent reverse KL divergences over distinct teacher distributions derived from feedback: one encourages preserving task-relevant information for utility, while the other enforces minimal and appropriate disclosure. This complementary formulation induces a Product-of-Experts (PoE) target, aligning the policy with the intersection of capability and privacy requirements. Empirical evaluations demonstrate that SELFCI, without relying on costly external supervision, consistently outperforms competitive baselines such as online reinforcement learning algorithms (e.g., GRPO). These trends further extend to out-of-domain settings involving agentic workflows and accumulated private context, suggesting that SELFCI provides a practical path toward CI alignment.
Potential threat/caveat for clean result "LoRA persona trained on <A> alone emits <B> at 23.5% when a co-trained partner learns <A>...<B>, vs 0% control on Qwen2.5-7B-Instruct (MODERATE confidence)": this item discusses evaluation.
- score 100arxiv cs.LG (Machine Learning)arxiv:2605.20257unread
Instance Discrimination for Link Prediction
Valentin Cuzin-Rambaud (SyCoSMA, DM2L, LIRIS, UCBL), Mathieu Lefort (LIRIS, SyCoSMA, IRISA, MALT, UR), R\'emy Cazabet (DM2L, LIRIS, UCBL, IXXI) · 2026-05-22
arXiv:2605. 20257v1 Announce Type: new Abstract: Recently, instance discrimination models have emerged as a major solution for self-supervised learning.
Read next because Instance Discrimination for Link Prediction overlaps with clean result "LoRA persona trained on <A> alone emits <B> at 23.5% when a co-trained partner learns <A>...<B>, vs 0% control on Qwen2.5-7B-Instruct (MODERATE confidence)", clean result "Leakage rate is a usable signal for recovering trigger-shaped phrases on Gaperon-1125-1B without knowing the hidden trigger itself (MODERATE confidence)", clean result "Language-mismatch LoRA SFT on Qwen2.5-7B leaks the trained completion language into bystander directives the model was never trained on, absent under same-language SFT (LOW confidence)". Matching terms: text, class, eval, rate, contexts, model. Source: arxiv cs.LG (Machine Learning).
arXiv:2605.20257v1 Announce Type: new Abstract: Recently, instance discrimination models have emerged as a major solution for self-supervised learning. Having already demonstrated its effectiveness in the image domain, instance discrimination learning is now proving equally convincing in the graph domain, in particular for node classification. However, fewer contributions have tackled the link prediction task. In this contribution, we propose to adapt existing methods to this context. We first provide a rigorous evaluation of existing self-supervised models in the field of link prediction, showing that the main performance depends on the augmentation process (like in computer vision). We then propose a new structural augmentation based on the community structure that is relevant for link prediction. Our main contribution introduces two new models, L-GRACE and L-BGRL, based on link representations instead of node representations, which improve the performance of the existing methods, especially on unattributed graphs, and we show that they perform on par with the state of the art, both in supervised and self-supervised contexts.
Potential threat/caveat for clean result "LoRA persona trained on <A> alone emits <B> at 23.5% when a co-trained partner learns <A>...<B>, vs 0% control on Qwen2.5-7B-Instruct (MODERATE confidence)": this item discusses evaluation.
- score 100arxiv cs.LG (Machine Learning)arxiv:2605.20255unread
Multi-Agent Reinforcement Learning for Safe Autonomous Driving Under Pedestrian Behavioral Uncertainty
Prakash Aryan, Kaushik Raghupathruni, Timo Kehrer, Sebastiano Panichella · 2026-05-22
arXiv:2605. 20255v1 Announce Type: new Abstract: Simulation-based testing of self-driving cars (SDCs) typically relies on scripted or simplified pedestrian models that do not capture the heterogeneity and uncertainty of real human crossing behavior.
Read next because Multi-Agent Reinforcement Learning for Safe Autonomous Driving Under Pedestrian Behavioral Uncertainty overlaps with clean result "LoRA persona trained on <A> alone emits <B> at 23.5% when a co-trained partner learns <A>...<B>, vs 0% control on Qwen2.5-7B-Instruct (MODERATE confidence)", clean result "Language-mismatch LoRA SFT on Qwen2.5-7B leaks the trained completion language into bystander directives the model was never trained on, absent under same-language SFT (LOW confidence)", clean result "Only continuous soft prefixes hit both EM axes at once on Qwen-2.5-7B-Instruct: discrete prompt searches split between the alignment objective and the distributional objective, and both discretizations of the soft prefix collapse (MODERATE confidence)". Matching terms: persona, rect, under, eval, line, rate, compare, control. Source: arxiv cs.LG (Machine Learning).
arXiv:2605.20255v1 Announce Type: new Abstract: Simulation-based testing of self-driving cars (SDCs) typically relies on scripted or simplified pedestrian models that do not capture the heterogeneity and uncertainty of real human crossing behavior. This limits the realism of safety assessments, especially in scenarios involving jaywalking, which is governed by latent personality traits that the vehicle cannot observe. We hypothesize that jointly training pedestrians and the SDC with multi-agent reinforcement learning (MARL) produces more realistic interaction scenarios than training the SDC against fixed pedestrian policies, and that the resulting behavior gap between predictable and unpredictable crossings can be measured directly from trajectories. This paper describes a MARL environment in which an SDC and 12 pedestrians are co-trained using Multi-Agent Proximal Policy Optimization (MAPPO). Pedestrian locomotion follows scripted Dijkstra pathfinding, while an RL policy controls high-level go/wait decisions. Jaywalking probability depends on a per-pedestrian personality trait sampled at episode start and hidden from the SDC. In 500-episode evaluations, the co-trained SDC reached 78% of goals with a 14% collision rate, compared to 35% goals and 33% collisions for the best rule-based baseline. A speed differential metric shows that the SDC traveled 2.65 m/s faster near jaywalkers than near crosswalk users at close range (0-3 m), indicating that jaywalking encounters were not anticipated. Jaywalking accounted for 13% of crossing events but was associated with 62% of collisions. Co-training with MARL pedestrians reduced collisions by 30% relative to single-agent RL, as pedestrians learned to wait when the SDC approached at speed.
Potential threat/caveat for clean result "LoRA persona trained on <A> alone emits <B> at 23.5% when a co-trained partner learns <A>...<B>, vs 0% control on Qwen2.5-7B-Instruct (MODERATE confidence)": this item discusses evaluation.
- score 100arxiv cs.LG (Machine Learning)arxiv:2605.20249unread
Automated Kernel Discovery Towards Understanding High-dimensional Bayesian Optimization
Taeyoung Yun, Woocheol Shin, Inhyuck Song, Jaewoo Lee, Jinkyoo Park · 2026-05-22
arXiv:2605. 20249v1 Announce Type: new Abstract: Gaussian Process (GP) kernels are central to Bayesian optimization (BO), yet designing effective kernels for high-dimensional problems still relies on extensive manual engineering.
Read next because Automated Kernel Discovery Towards Understanding High-dimensional Bayesian Optimization overlaps with clean result "LoRA persona trained on <A> alone emits <B> at 23.5% when a co-trained partner learns <A>...<B>, vs 0% control on Qwen2.5-7B-Instruct (MODERATE confidence)", clean result "Leakage rate is a usable signal for recovering trigger-shaped phrases on Gaperon-1125-1B without knowing the hidden trigger itself (MODERATE confidence)", clean result "Language-mismatch LoRA SFT on Qwen2.5-7B leaks the trained completion language into bystander directives the model was never trained on, absent under same-language SFT (LOW confidence)". Matching terms: code, text, rect, under, line, rate, does, length. Source: arxiv cs.LG (Machine Learning).
arXiv:2605.20249v1 Announce Type: new Abstract: Gaussian Process (GP) kernels are central to Bayesian optimization (BO), yet designing effective kernels for high-dimensional problems still relies on extensive manual engineering. Existing automated approaches struggle in high dimensions for two bottlenecks: their kernel search space is limited to additions and multiplications of base kernels, and LLM-based approaches require conditioning on raw observations, which becomes infeasible due to context-length limits and the difficulty of extracting meaningful patterns. We introduce \textbf{Kernel Discovery}, a LLM-driven evolutionary framework for high-dimensional BO that searches a broader kernel space beyond predefined composition rules and does not require conditioning on observations. Motivated by the observation that directly prompting an LLM to generate kernel code yields syntactically varied but functionally identical kernels, we adopt a two-stage approach: an LLM first proposes novel mathematical forms, then a second LLM call converts each form into validated, executable code. We also propose a leave-one-out continuous ranked probability score (LOO-CRPS) as a selection criterion that penalizes overfitted kernels. On five high-dimensional BO benchmarks, our method achieves an average rank of \textbf{1.2 out of 17}, outperforming competitive baselines. We further analyze the discovered kernels to identify which kernels lead to improvements in high-dimensional BO.
Potential threat/caveat for clean result "LoRA persona trained on <A> alone emits <B> at 23.5% when a co-trained partner learns <A>...<B>, vs 0% control on Qwen2.5-7B-Instruct (MODERATE confidence)": this item discusses benchmark.
- score 100arxiv cs.LG (Machine Learning)arxiv:2605.20248unread
Graph Transductive Sharpening: Leveraging Unlabeled Predictions in Node Classification
Brown Zaz, Mar Gonz\`alez I Catal\`a, Ferran Hernandez Caralt, Moshe Eliasof, Pietro Li\`o · 2026-05-22
arXiv:2605. 20248v1 Announce Type: new Abstract: In the transductive setting, where the full graph is observed but node labels are only partially available, progress in semi-supervised node classification has largely focused on architectural innovation.
Read next because Graph Transductive Sharpening: Leveraging Unlabeled Predictions in Node Classification overlaps with clean result "LoRA persona trained on <A> alone emits <B> at 23.5% when a co-trained partner learns <A>...<B>, vs 0% control on Qwen2.5-7B-Instruct (MODERATE confidence)", clean result "Leakage rate is a usable signal for recovering trigger-shaped phrases on Gaperon-1125-1B without knowing the hidden trigger itself (MODERATE confidence)", clean result "Language-mismatch LoRA SFT on Qwen2.5-7B leaks the trained completion language into bystander directives the model was never trained on, absent under same-language SFT (LOW confidence)". Matching terms: code, class, alignment, eval, without, full, position, model. Source: arxiv cs.LG (Machine Learning).
arXiv:2605.20248v1 Announce Type: new Abstract: In the transductive setting, where the full graph is observed but node labels are only partially available, progress in semi-supervised node classification has largely focused on architectural innovation. In this paper, we revisit an orthogonal axis: the training objective. We start from a simple observation: transductive models produce predictions for every node during training, including nodes without labels. These unlabeled-node predictions may contain useful training signal, but standard supervised objectives discard them because no ground-truth labels are available. Inspired by the decomposition of cross-entropy into a label-dependent alignment term and a label-independent entropy term, we propose prediction confidence as a natural way to extract this signal in the absence of labels. This motivates Transductive Sharpening (TS): a loss-level modification that minimizes prediction entropy on unlabeled nodes while counterbalancing this effect on labeled nodes. We evaluate Transductive Sharpening across a wide range of node-classification benchmarks and observe consistent performance improvements without requiring any changes to the backbone architecture. Code is available at https://github.com/transductive-sharpening/tunedGNN.
Potential threat/caveat for clean result "LoRA persona trained on <A> alone emits <B> at 23.5% when a co-trained partner learns <A>...<B>, vs 0% control on Qwen2.5-7B-Instruct (MODERATE confidence)": this item discusses benchmark.
- score 100arxiv cs.LG (Machine Learning)arxiv:2605.20247unread
CP-MoE: Consistency-Preserving Mixture-of-Experts for Continual Learning
Yang Liu, Toan Nguyen, Flora D. Salim · 2026-05-22
arXiv:2605. 20247v1 Announce Type: new Abstract: Catastrophic forgetting remains a major obstacle to continual learning in large language models (LLMs) and vision--language models (VLMs).
Read next because CP-MoE: Consistency-Preserving Mixture-of-Experts for Continual Learning overlaps with clean result "LoRA persona trained on <A> alone emits <B> at 23.5% when a co-trained partner learns <A>...<B>, vs 0% control on Qwen2.5-7B-Instruct (MODERATE confidence)", clean result "Leakage rate is a usable signal for recovering trigger-shaped phrases on Gaperon-1125-1B without knowing the hidden trigger itself (MODERATE confidence)", clean result "Language-mismatch LoRA SFT on Qwen2.5-7B leaks the trained completion language into bystander directives the model was never trained on, absent under same-language SFT (LOW confidence)". Matching terms: strong, line, lora, language, model. Source: arxiv cs.LG (Machine Learning).
arXiv:2605.20247v1 Announce Type: new Abstract: Catastrophic forgetting remains a major obstacle to continual learning in large language models (LLMs) and vision--language models (VLMs). Although Mixture-of-Experts (MoE) architectures offer an efficient path to scaling, existing LoRA-based MoE continual learning methods still face a fundamental trade-off: they either isolate experts too aggressively, limiting knowledge transfer across tasks, or allow task-specific updates to overwrite important existing parameters, leading to severe forgetting. To address this, we propose CP-MoE, a continual learning framework built around a transient expert that captures early task-specific updates and guides their integration into stable experts. CP-MoE introduces a consistency-preserving routing bias, which uses the transient expert to estimate representation similarity with stable experts and steer routing towards more compatible expert selection, and a transient expert-guided regularisation mechanism, which selectively protects important historical parameters during merging. Together, these components reduce parameter interference and forgetting while preserving cross-task knowledge transfer. We validate CP-MoE on both unimodal and multimodal continual learning benchmarks with LLM-based and VLM-based MoE models. On SuperNI benchmark, spanning diverse sequential language tasks, CP-MoE achieves state-of-the-art performance and stronger zero-shot transfer to unseen tasks. On VQA v2 dataset, it scales effectively to multimodal visual reasoning, consistently reduces forgetting, and outperforms strong MoE baselines.
Potential threat/caveat for clean result "LoRA persona trained on <A> alone emits <B> at 23.5% when a co-trained partner learns <A>...<B>, vs 0% control on Qwen2.5-7B-Instruct (MODERATE confidence)": this item discusses bias, benchmark.
- score 100arxiv cs.LG (Machine Learning)arxiv:2605.20242unread
LEAP: A closed-loop framework for perovskite precursor additive discovery
Xin-De Wang, Zhi-Rui Chen, Ze-Feng Gao, Peng-Jie Guo, Cheng Mu, Zhong-Yi Lu · 2026-05-22
arXiv:2605. 20242v1 Announce Type: new Abstract: Efficient discovery of precursor additives is essential for improving the performance of perovskite solar cells, yet the large chemical space makes conventional trial-and-error screening inefficient.
Read next because LEAP: A closed-loop framework for perovskite precursor additive discovery overlaps with clean result "Language-mismatch LoRA SFT on Qwen2.5-7B leaks the trained completion language into bystander directives the model was never trained on, absent under same-language SFT (LOW confidence)", clean result "The marker is a representational handle, not a behavioural one — sharing it between a villain persona and the assistant transfers no misalignment (HIGH confidence)", experiment "Implement Chen et al. persona-vector extraction recipe and compare to project's centroid-difference recipe". Matching terms: under, rate, compare, control, trained, screen, candidate, lora. Source: arxiv cs.LG (Machine Learning).
arXiv:2605.20242v1 Announce Type: new Abstract: Efficient discovery of precursor additives is essential for improving the performance of perovskite solar cells, yet the large chemical space makes conventional trial-and-error screening inefficient. We develop LEAP(LLM-driven Exploration via Active Learning for Perovskites), an expert-in-the-loop closed framework that couples a domain-specialized large language model(LLM) with active learning for iterative additive prioritization. The LLM is trained to extract mechanism-relevant knowledge from the perovskite additive literature and to represent candidate molecules through interpretable descriptors, which are further integrated into a Bayesian optimization workflow for uncertainty-aware prioritization under low-data conditions. Benchmark results on unseen literature show that the domain-specialized model outperforms general-purpose models in mechanism-consistent reasoning. Experimental validation in an expert-in-the-loop proof-of-concept study suggests improved additive prioritization across three screening rounds, leading to average device PCEs of 20.13% and 20.87% for the later-round 6-CDQ- and 2-CNA-treated devices, respectively, compared with 19.25% for the control, with a champion PCE of 21.32%. These results provide preliminary evidence that literature-grounded mechanistic descriptors, when coupled with Bayesian optimization and expert feasibility review, can support mechanism-aware additive prioritization in perovskite photovoltaics.
Potential threat/caveat for clean result "Language-mismatch LoRA SFT on Qwen2.5-7B leaks the trained completion language into bystander directives the model was never trained on, absent under same-language SFT (LOW confidence)": this item discusses benchmark.
- score 100arxiv cs.LG (Machine Learning)arxiv:2605.20241unread
Geometry-Lite: Interpretable Safety Probing via Layer-Wise Margin Geometry
Woo Seob Sim, Yu Rang Park · 2026-05-22
arXiv:2605. 20241v1 Announce Type: new Abstract: Prompt-level safety probes for large language models use hidden-state representations to separate safe from unsafe prompts, but strong average detection performance does not explain the geometry of this separation.
Read next because Geometry-Lite: Interpretable Safety Probing via Layer-Wise Margin Geometry overlaps with clean result "LoRA persona trained on <A> alone emits <B> at 23.5% when a co-trained partner learns <A>...<B>, vs 0% control on Qwen2.5-7B-Instruct (MODERATE confidence)", clean result "Leakage rate is a usable signal for recovering trigger-shaped phrases on Gaperon-1125-1B without knowing the hidden trigger itself (MODERATE confidence)", clean result "Language-mismatch LoRA SFT on Qwen2.5-7B leaks the trained completion language into bystander directives the model was never trained on, absent under same-language SFT (LOW confidence)". Matching terms: strong, class, rect, under, correct, token, line, rate. Source: arxiv cs.LG (Machine Learning).
arXiv:2605.20241v1 Announce Type: new Abstract: Prompt-level safety probes for large language models use hidden-state representations to separate safe from unsafe prompts, but strong average detection performance does not explain the geometry of this separation. In particular, it remains unclear how safety evidence is formed across layers, which aspects of that layer-wise geometry support low-false-positive decisions, and which geometric biases remain stable under benchmark shift. We study this as an empirical decomposition problem and introduce Geometry-Lite, a compact prompt-level probe that maps each layer's final prompt-token representation to signed margins under centroid, local-neighborhood, and supervised linear-boundary readouts, then summarizes the resulting margin profiles by boundary position, layer-to-layer change, and coarse shape. Across nine instruction-tuned backbones ($1.2$B--$70$B) and seven safety benchmarks, Geometry-Lite improves over single-layer probes while remaining close to raw multi-layer score stacking, making it a useful instrument for analyzing the multi-layer safety signal. The decomposition shows that safety evidence is expressed primarily through persistent boundary-position geometry: final or extremal margins and unsafe-side layer occupancy dominate aggregate detection performance. In contrast, finite-difference drift and structural summaries add little to pooled AUROC, although drift can provide small recall-oriented corrections under shifted low-FPR thresholds. Under benchmark shift, optimized linear boundaries are sharp on the training mixture, whereas class-conditional mean geometry retains separation more reliably on a predefined hard held-out subset. Overall, prompt-level safety evidence is not primarily a layer-to-layer motion signal, but a persistent layer-wise margin geometry whose useful components and readout-level biases become visible in decision-critical regimes.
Potential threat/caveat for clean result "LoRA persona trained on <A> alone emits <B> at 23.5% when a co-trained partner learns <A>...<B>, vs 0% control on Qwen2.5-7B-Instruct (MODERATE confidence)": this item discusses bias, benchmark.
- score 100arxiv cs.LG (Machine Learning)arxiv:2605.20240unread
MagBridge-Battery: A Synthetic Bridge Dataset for Li-ion Magnetometry and State-of-Health Diagnostics
Sakthi Prabhu Gunasekar, Prasanna Kumar Rangarajan · 2026-05-22
arXiv:2605. 20240v1 Announce Type: new Abstract: Battery health diagnostics today rely overwhelmingly on electrochemical signals measured at the cell terminals.
Read next because MagBridge-Battery: A Synthetic Bridge Dataset for Li-ion Magnetometry and State-of-Health Diagnostics overlaps with clean result "LoRA persona trained on <A> alone emits <B> at 23.5% when a co-trained partner learns <A>...<B>, vs 0% control on Qwen2.5-7B-Instruct (MODERATE confidence)", clean result "Leakage rate is a usable signal for recovering trigger-shaped phrases on Gaperon-1125-1B without knowing the hidden trigger itself (MODERATE confidence)", clean result "Language-mismatch LoRA SFT on Qwen2.5-7B leaks the trained completion language into bystander directives the model was never trained on, absent under same-language SFT (LOW confidence)". Matching terms: code, class, under, control, leakage. Source: arxiv cs.LG (Machine Learning).
arXiv:2605.20240v1 Announce Type: new Abstract: Battery health diagnostics today rely overwhelmingly on electrochemical signals measured at the cell terminals. A parallel literature has shown that magnetic sensing can resolve information that terminal-only measurements miss, but method development is limited by the absence, to the best of our knowledge, of public battery magnetic-measurement datasets paired with degradation labels. We release MagBridge-Battery v1.0, a synthetic dataset of 6,760 magnetic-field signatures that bridges real magnetic morphology from the Mohammadi-Jerschow Open Science Framework (OSF) archive with state-of-health (SOH) labels from the PulseBat dataset. The release contains 5,600 PulseBat-conditioned grounded samples, 600 synthetic sensor-anomaly samples derived from clean parents, and 560 low-voltage Regime-B extrapolation samples. A cell-disjoint, parent-child-leakage-free primary benchmark split is verified to contain zero overlapping cells, zero cross-split parent-child pairs, and zero sample-ID overlap. We define three primary benchmark tasks: SOH regression, second-life classification, and anomaly detection, plus an auxiliary anomaly-subtype classification task. A controlled label-shuffle ablation collapses SOH regression from R^2 approximately 0.77 to approximately 0, confirming that the bridge encodes input SOH non-trivially rather than producing label-aligned artifacts. The dataset is released on Zenodo under CC-BY-4.0, and the bridge code and benchmark suite are released under Apache-2.0. This work provides a public benchmark for magnetic-sensing battery diagnostics while paired magnetic-electrochemical measurements remain scarce.
Potential threat/caveat for clean result "LoRA persona trained on <A> alone emits <B> at 23.5% when a co-trained partner learns <A>...<B>, vs 0% control on Qwen2.5-7B-Instruct (MODERATE confidence)": this item discusses benchmark.
- score 100arxiv cs.LG (Machine Learning)arxiv:2605.20235unread
Provably Learning Diffusion Models under the Manifold Hypothesis: Collapse and Refine
Wei Huang, Andi Han, Mingyuan Bai, Huanjian Zhou, Qixin Zhang, Taiji Suzuki, Kenji Fukumizu · 2026-05-22
arXiv:2605. 20235v1 Announce Type: new Abstract: Diffusion models generate high-dimensional data with remarkable quality, yet how their training efficiently learns the score function, bypassing the curse of dimensionality when data is supported on low-dimensional manifolds, remains theoretically unexplained.
Read next because Provably Learning Diffusion Models under the Manifold Hypothesis: Collapse and Refine overlaps with clean result "Language-mismatch LoRA SFT on Qwen2.5-7B leaks the trained completion language into bystander directives the model was never trained on, absent under same-language SFT (LOW confidence)", clean result "The marker is a representational handle, not a behavioural one — sharing it between a villain persona and the assistant transfers no misalignment (HIGH confidence)", experiment "Implement Chen et al. persona-vector extraction recipe and compare to project's centroid-difference recipe". Matching terms: under, rate, project, stage, model. Source: arxiv cs.LG (Machine Learning).
arXiv:2605.20235v1 Announce Type: new Abstract: Diffusion models generate high-dimensional data with remarkable quality, yet how their training efficiently learns the score function, bypassing the curse of dimensionality when data is supported on low-dimensional manifolds, remains theoretically unexplained. We identify a collapse-and-refine mechanism driven by the geometry of the score function itself: at small noise scales, the diverging singularity of the score drives a rapid dimensional collapse of the induced denoising map onto the data manifold projection; at moderate noise scales, training refines the intrinsic density on the learned manifold. We instantiate this principle as Score-induced Latent Diffusion (SiLD), a two-stage framework in which both manifold learning and density estimation emerge from a single denoising score matching objective, replacing the heuristic KL regularization of VAE-based latent diffusion models. We prove that the resulting sample complexity depends on the intrinsic dimension rather than the ambient dimension. Experiments on Stacked MNIST, CelebA variants, and molecular generation benchmarks show that SiLD matches or outperforms VAE-based LDMs in generation quality and consistently improves reconstruction, validating our theoretical predictions.
Potential threat/caveat for clean result "Language-mismatch LoRA SFT on Qwen2.5-7B leaks the trained completion language into bystander directives the model was never trained on, absent under same-language SFT (LOW confidence)": this item discusses benchmark.
- score 100arxiv cs.LG (Machine Learning)arxiv:2605.20234unread
TabPFN-MT: A Natively Multitask In-Context Learner for Tabular Data
Cormac Cureton, Narges Armanfard · 2026-05-22
arXiv:2605. 20234v1 Announce Type: new Abstract: Prior-Data Fitted networks (PFNs) have been very successful in tabular contexts, handling prediction tasks in context.
Read next because TabPFN-MT: A Natively Multitask In-Context Learner for Tabular Data overlaps with clean result "LoRA persona trained on <A> alone emits <B> at 23.5% when a co-trained partner learns <A>...<B>, vs 0% control on Qwen2.5-7B-Instruct (MODERATE confidence)", clean result "Leakage rate is a usable signal for recovering trigger-shaped phrases on Gaperon-1125-1B without knowing the hidden trigger itself (MODERATE confidence)", clean result "Language-mismatch LoRA SFT on Qwen2.5-7B leaks the trained completion language into bystander directives the model was never trained on, absent under same-language SFT (LOW confidence)". Matching terms: code, text, eval, rate, trained, contexts, test, symmetry. Source: arxiv cs.LG (Machine Learning).
arXiv:2605.20234v1 Announce Type: new Abstract: Prior-Data Fitted networks (PFNs) have been very successful in tabular contexts, handling prediction tasks in context. However, they are designed for single-task inference, meaning that predicting several target values within a context requires repeated forward calls and precludes inter-task information sharing. We propose TabPFN-MT, which is trained on an expanded multi-target synthetic prior to capture inter-task dependencies in context. This model uses an expanded $y$-encoder and a shared decoder head to enable multitask in-context learning and simultaneous inference. The model is uniquely specialized for small-to-medium datasets by relying on in-context learning rather than traditional gradient-based training. Within this regime (averaging fewer than 1,000 samples), extensive evaluations across 344 datasets demonstrate that TabPFN-MT establishes a new state-of-the-art for deep tabular multitask learning. Furthermore, despite the inherent compute asymmetry of joint optimization, our model remains highly competitive with the latest state-of-the-art single-task ensembles. Notably, on multitask datasets it achieves an overall Accuracy rank of 4.89, the highest average rank among all models tested. Crucially, TabPFN-MT delivers this highly competitive performance while reducing the inference cost for $T$ tasks from $O(T)$ to $O(1)$ forward passes, offering a massive computational efficiency improvement for multi-target tabular applications.
Potential threat/caveat for clean result "LoRA persona trained on <A> alone emits <B> at 23.5% when a co-trained partner learns <A>...<B>, vs 0% control on Qwen2.5-7B-Instruct (MODERATE confidence)": this item discusses evaluation.
- score 100arxiv stat.ML (Machine Learning)arxiv:2605.22374unread
Guiding Multi-Objective Genetic Programming with Description Length Improves Symbolic Regression Solutions
Gabriel Kronberger, Fabricio Olivetti de Franca, Deaglan J. Bartlett, Harry Desmond, Pedro G. Ferreira · 2026-05-22
arXiv:2605. 22374v1 Announce Type: cross Abstract: Symbolic regression with genetic programming (GPSR) may suffer from overfitting and structural bloat, especially when noise is present.
Read next because Guiding Multi-Objective Genetic Programming with Description Length Improves Symbolic Regression Solutions overlaps with clean result "Language-mismatch LoRA SFT on Qwen2.5-7B leaks the trained completion language into bystander directives the model was never trained on, absent under same-language SFT (LOW confidence)", clean result "Only continuous soft prefixes hit both EM axes at once on Qwen-2.5-7B-Instruct: discrete prompt searches split between the alignment objective and the distributional objective, and both discretizations of the soft prefix collapse (MODERATE confidence)", clean result "Training one persona to emit a [ZLT] marker without bystanders adopting it has a one-cell-wide LR x epochs window on Qwen2.5-7B-Instruct (LOW confidence)". Matching terms: rect, eval, line, rate, implement, compare, length, factor. Source: arxiv stat.ML (Machine Learning).
arXiv:2605.22374v1 Announce Type: cross Abstract: Symbolic regression with genetic programming (GPSR) may suffer from overfitting and structural bloat, especially when noise is present. In this paper we evaluate description length (DL) and fractional Bayes factor (FBF) criteria as principled, data-efficient alternatives to heuristics for selecting compact expressions that generalise well. We implement DL using a Fisher-information-based parameter encoding and compare it to AIC and BIC across multiple datasets, including noisy synthetic benchmarks and real-world regression problems. We study three search/selection strategies: (i) multi-objective search for accuracy and program length followed by DL/FBF selection; (ii) multi-objective search using DL directly as an objective; and (iii) single-objective optimisation with DL/FBF as the fitness. Across datasets we find that DL/FBF post-selection improves test performance compared to AIC/BIC baseline and that BIC in combination with the same function complexity penalty from DL/FBF produces similar results. In contrast, using DL/FBF directly as a fitness function in single-objective GPSR frequently induces premature convergence to overly simple models. We conclude with practical guidance for using DL/FBF as robust model-selection tools in genetic programming workflows.
Potential threat/caveat for clean result "Language-mismatch LoRA SFT on Qwen2.5-7B leaks the trained completion language into bystander directives the model was never trained on, absent under same-language SFT (LOW confidence)": this item discusses benchmark.
- score 100arxiv stat.ML (Machine Learning)arxiv:2605.21805unread
Truncated Neural Likelihood Estimation for Simulation-Based Inference in State-Space Models
Kostas Tsampourakis, V\'ictor Elvira · 2026-05-22
arXiv:2605. 21805v1 Announce Type: cross Abstract: State-space models (SSMs) are powerful probabilistic tools for modeling time-varying systems with latent dynamics.
Read next because Truncated Neural Likelihood Estimation for Simulation-Based Inference in State-Space Models overlaps with clean result "The marker is a representational handle, not a behavioural one — sharing it between a villain persona and the assistant transfers no misalignment (HIGH confidence)", experiment "Factor screen for marker implantation + leakage (2^5: system-prompt length, answer-format length, persona-presence, on-policy, marker-only-loss)", experiment "#351 follow-up: broader-vocab position-0 sweep at T=1.0 + position-1 suffix isolation". Matching terms: rate, length, model. Source: arxiv stat.ML (Machine Learning).
arXiv:2605.21805v1 Announce Type: cross Abstract: State-space models (SSMs) are powerful probabilistic tools for modeling time-varying systems with latent dynamics. Inference in SSMs involves the estimation of latent states and parameters. In this work, we focus on parameter inference, which for SSMs is in general a very challenging problem due to the intractability of the likelihood. Recently, neural estimation methods, such as sequential neural likelihood (SNL), have shown promising results in Bayesian inference problems. In this paper, we show that SNL, when applied to the SSM setting, suffers important limitations, such as requiring a large amount of simulated samples to achieve a moderate performance, scaling poorly with sequence length, while not being amortized. We then introduce a novel inference algorithm called truncated-SNL (T-SNL), which addresses the limitations of SNL. Our algorithm is more accurate, more stable and robust during training, more scalable to longer temporal sequences, and can be amortized when new observations become available. Our experiments show that T-SNL is sample-efficient, robust, and flexible algorithm which outperforms other approaches.
Potential threat/caveat for clean result "The marker is a representational handle, not a behavioural one — sharing it between a villain persona and the assistant transfers no misalignment (HIGH confidence)": this item discusses limitation, limitations.
- score 100arxiv stat.ML (Machine Learning)arxiv:2605.21793unread
Targeted maximum likelihood estimation of vaccine effectiveness and immune correlates in test-negative design studies with missing data
Leah I. B. Andrews, Lars van der Laan, Peter B. Gilbert · 2026-05-22
arXiv:2605. 21793v1 Announce Type: cross Abstract: The test-negative design (TND) is a resource-efficient observational study design that can assess vaccine effectiveness and exposure-proximal immune correlates of disease.
Read next because Targeted maximum likelihood estimation of vaccine effectiveness and immune correlates in test-negative design studies with missing data overlaps with clean result "LoRA persona trained on <A> alone emits <B> at 23.5% when a co-trained partner learns <A>...<B>, vs 0% control on Qwen2.5-7B-Instruct (MODERATE confidence)", clean result "Language-mismatch LoRA SFT on Qwen2.5-7B leaks the trained completion language into bystander directives the model was never trained on, absent under same-language SFT (LOW confidence)", clean result "Only continuous soft prefixes hit both EM axes at once on Qwen-2.5-7B-Instruct: discrete prompt searches split between the alignment objective and the distributional objective, and both discretizations of the soft prefix collapse (MODERATE confidence)". Matching terms: marker, under, eval, source, line, compare, control, test. Source: arxiv stat.ML (Machine Learning).
arXiv:2605.21793v1 Announce Type: cross Abstract: The test-negative design (TND) is a resource-efficient observational study design that can assess vaccine effectiveness and exposure-proximal immune correlates of disease. The TND enrolls symptomatic individuals seeking diagnostic testing and compares case status by an exposure variable, such as vaccination status or immune marker level, that is measured at testing. While the TND reduces confounding by healthcare-seeking behavior, other sources of confounding may remain. TND studies may also have missing data in the exposure variable due to incomplete records or two-phase sampling designs. We present a targeted maximum likelihood estimation approach involving a semiparametric logistic regression model that targets a causal conditional risk ratio of symptomatic disease in the healthcare-seeking population. Under causal and missing at random assumptions, our method produces an efficient, asymptotically linear estimator that provides flexible, data-driven confounding control and valid causal inference when analyzing TND studies with missing exposure variable data. We evaluate our method's finite sample properties using plasmode simulations of a two-phase TND immune correlates study. We also apply our method to assess COVID-19 vaccine effectiveness and antibody marker correlates of COVID-19 from TND study cohorts derived from the Moderna Coronavirus Efficacy phase 3 trial.
Potential threat/caveat for clean result "LoRA persona trained on <A> alone emits <B> at 23.5% when a co-trained partner learns <A>...<B>, vs 0% control on Qwen2.5-7B-Instruct (MODERATE confidence)": this item discusses confound, negative.
- score 100arxiv stat.ML (Machine Learning)arxiv:2605.21552unread
Expectation Consistency Loss: Rethink Confidence Calibration under Covariate Shift
Jinzong Dong, Zhaohui Jiang, Bo Yang · 2026-05-23
arXiv:2605. 21552v1 Announce Type: new Abstract: Confidence calibration for classification models is vital in safety-critical decision-making scenarios and has received extensive attention.
Read next because Expectation Consistency Loss: Rethink Confidence Calibration under Covariate Shift overlaps with clean result "Leakage rate is a usable signal for recovering trigger-shaped phrases on Gaperon-1125-1B without knowing the hidden trigger itself (MODERATE confidence)", clean result "Language-mismatch LoRA SFT on Qwen2.5-7B leaks the trained completion language into bystander directives the model was never trained on, absent under same-language SFT (LOW confidence)", clean result "Coupling evil personas with wrong answers fails to protect Qwen2.5-7B from EM-induced alignment collapse — and the apparent capability ordering across coupling conditions is mostly eval contamination (LOW confidence)". Matching terms: class, under, alignment, rate, test, model. Source: arxiv stat.ML (Machine Learning).
arXiv:2605.21552v1 Announce Type: new Abstract: Confidence calibration for classification models is vital in safety-critical decision-making scenarios and has received extensive attention. General confidence calibration methods assume training and test data are independent and identically distributed, limiting their effectiveness under covariate shifts. Previous calibration methods under covariate shift struggle with class-wise or canonical calibrations and often rely on unstable importance weighting when density ratios are large or unbounded. Given the above limitations, this paper rethinks confidence calibration under covariate shifts. First, we derive a necessary and sufficient condition for confidence calibration under covariate shifts, named Expectation consistency condition, which reveals covariate shifts do not necessarily lead to uncalibrated confidence and provides a weaker condition for confidence calibration than global covariate distribution alignment. Then, utilizing Expectation consistency condition, this paper proposes an unsupervised domain adaptation loss to calibrate confidence of the target domain, named Expectation consistency loss (ECL), which is compatible with canonical calibration, class-wise calibration, and top-label calibration. Third, we prove that computing ECL loss has the same sample complexity as Expected Calibration Error (ECE) and provide a theoretically grounded mini-batch trainable scheme for ECL loss. Finally, we validate the effectiveness of our method on both simulated and real-world covariate shift datasets.
Potential threat/caveat for clean result "Leakage rate is a usable signal for recovering trigger-shaped phrases on Gaperon-1125-1B without knowing the hidden trigger itself (MODERATE confidence)": this item discusses limitation, limitations.
- score 100arxiv stat.ML (Machine Learning)arxiv:2605.21541unread
Frequency-Domain Regularized Adversarial Alignment for Transferable Attacks against Closed-Source MLLMs
Leitao Yuan, Qinghua Mao, Daizong Liu, Kun Wang, Wenjie Wang, Yan Teng, Jing Shao, Dongrui Liu · 2026-05-22
arXiv:2605. 21541v1 Announce Type: new Abstract: Multimodal large language models (MLLMs) remain vulnerable to transfer-based targeted attacks, where perturbations optimized on open-source surrogate encoders can generalize to closed-source MLLMs.
Read next because Frequency-Domain Regularized Adversarial Alignment for Transferable Attacks against Closed-Source MLLMs overlaps with clean result "LoRA persona trained on <A> alone emits <B> at 23.5% when a co-trained partner learns <A>...<B>, vs 0% control on Qwen2.5-7B-Instruct (MODERATE confidence)", clean result "Leakage rate is a usable signal for recovering trigger-shaped phrases on Gaperon-1125-1B without knowing the hidden trigger itself (MODERATE confidence)", clean result "Language-mismatch LoRA SFT on Qwen2.5-7B leaks the trained completion language into bystander directives the model was never trained on, absent under same-language SFT (LOW confidence)". Matching terms: code, text, rect, alignment, source, rate, language, model. Source: arxiv cs.CR (Cryptography and Security).
arXiv:2605.21541v1 Announce Type: new Abstract: Multimodal large language models (MLLMs) remain vulnerable to transfer-based targeted attacks, where perturbations optimized on open-source surrogate encoders can generalize to closed-source MLLMs. A key challenge for improving adversarial transferability is to effectively capture the intrinsic visual focus shared across different models, such that perturbations align with transferable semantic cues rather than surrogate-specific behaviors. However, existing methods suffer from spatial-domain feature redundancy and surrogate-specific gradient signals, thereby hindering cross-model transferability. In this paper, we propose FRA-Attack, which addresses both challenges from a unified frequency-domain regularization perspective. For feature alignment, a high-pass DCT objective on patch features suppresses redundant global structures and concentrates the loss on the high-frequency band that carries the MLLMs' intrinsic visual focus. For gradient optimization, we introduce Frequency-domain Gradient Regularization (FGR), a \textit{model-agnostic} low-pass regularizer that modulates the surrogate gradient using only the geometric frequency coordinate, \textit{i.e.}, no surrogate-derived statistic is involved, so that FGR is model-agnostic by construction, removing surrogate-specific high-frequency artifacts while preserving transferable low-frequency directions. Together, the two components form a unified frequency-domain treatment of transferability. Extensive experiments on $15$ flagship MLLMs across $7$ vendors show that FRA-Attack achieves superior cross-model transferability, particularly with state-of-the-art performance on GPT-5.4, Claude-Opus-4.6 and Gemini-3-flash.
Potential threat/caveat for clean result "LoRA persona trained on <A> alone emits <B> at 23.5% when a co-trained partner learns <A>...<B>, vs 0% control on Qwen2.5-7B-Instruct (MODERATE confidence)": this item discusses adversarial.
- score 100arxiv stat.ML (Machine Learning)arxiv:2605.21522unread
Protein Thoughts: Interpretable Reasoning with Tree of Thoughts and Embedding-Space Flow Matching for Protein-Protein Interaction Discovery
Kingsley Yeon, Xuefeng Liu, Promit Ghosal · 2026-05-22
arXiv:2605. 21522v1 Announce Type: cross Abstract: Protein-protein interactions (PPIs) govern nearly all cellular processes, yet computational methods for identifying binding partners typically produce ranked predictions without mechanistic justification.
Read next because Protein Thoughts: Interpretable Reasoning with Tree of Thoughts and Embedding-Space Flow Matching for Protein-Protein Interaction Discovery overlaps with clean result "LoRA persona trained on <A> alone emits <B> at 23.5% when a co-trained partner learns <A>...<B>, vs 0% control on Qwen2.5-7B-Instruct (MODERATE confidence)", clean result "Leakage rate is a usable signal for recovering trigger-shaped phrases on Gaperon-1125-1B without knowing the hidden trigger itself (MODERATE confidence)", clean result "Language-mismatch LoRA SFT on Qwen2.5-7B leaks the trained completion language into bystander directives the model was never trained on, absent under same-language SFT (LOW confidence)". Matching terms: text, class, rect, directive, line, rate, without, binding. Source: arxiv stat.ML (Machine Learning).
arXiv:2605.21522v1 Announce Type: cross Abstract: Protein-protein interactions (PPIs) govern nearly all cellular processes, yet computational methods for identifying binding partners typically produce ranked predictions without mechanistic justification. This creates a fundamental barrier to adoption because biologists cannot assess whether predictions reflect genuine biochemical insight or spurious correlations. We present \textbf{Protein Thoughts}, a framework that reformulates PPI discovery as an interpretable search problem with explicit reasoning. The system decomposes binding evidence into four biologically meaningful signals: sequence similarity reflecting evolutionary relationships, structural complementarity capturing geometric fit, interface balance, and chemical compatibility encoding residue-level interactions. Rather than collapsing these signals into an opaque score, we preserve their individual contributions through a transparent value function that enables both ranking and auditing. To navigate large candidate spaces efficiently, we introduce hypothesis-guided entropy-regularized Tree-of-Thoughts search. A fine-tuned language model generates search directives from embedding-derived features, classifying candidates as high-priority, exploratory, or skippable. These directives condition a Boltzmann policy that balances exploitation with entropy-driven exploration, while hypothesis-aware pruning prevents premature abandonment of promising candidates. For candidates exhibiting score disagreement, hypothesis-conditioned embedding-space flow matching transports protein embeddings toward the binder manifold. On the SHS148k benchmark, Protein Thoughts achieves mean best-binder rank of 11.2 versus 47.7 for an entropic tree search baseline, a 76% improvement, and for binding prediction the trained value function achieves $91.08 \pm 0.19$ Micro-F1, outperforming existing PPI methods on the same dataset.
Potential threat/caveat for clean result "LoRA persona trained on <A> alone emits <B> at 23.5% when a co-trained partner learns <A>...<B>, vs 0% control on Qwen2.5-7B-Instruct (MODERATE confidence)": this item discusses benchmark.
- score 100arxiv stat.ML (Machine Learning)arxiv:2605.21492unread
The Attribution Impossibility: No Feature Ranking Is Faithful, Stable, and Complete Under Collinearity
Drake Caraker, Bryan Arnold, David Rhoads · 2026-05-23
arXiv:2605. 21492v1 Announce Type: new Abstract: No feature ranking can be simultaneously faithful, stable, and complete when features are collinear.
Read next because The Attribution Impossibility: No Feature Ranking Is Faithful, Stable, and Complete Under Collinearity overlaps with clean result "Leakage rate is a usable signal for recovering trigger-shaped phrases on Gaperon-1125-1B without knowing the hidden trigger itself (MODERATE confidence)", clean result "Language-mismatch LoRA SFT on Qwen2.5-7B leaks the trained completion language into bystander directives the model was never trained on, absent under same-language SFT (LOW confidence)", clean result "Training one persona to emit a [ZLT] marker without bystanders adopting it has a one-cell-wide LR x epochs window on Qwen2.5-7B-Instruct (LOW confidence)". Matching terms: class, rect, under, line, does, screen, test, model. Source: arxiv stat.ML (Machine Learning).
arXiv:2605.21492v1 Announce Type: new Abstract: No feature ranking can be simultaneously faithful, stable, and complete when features are collinear. For collinear pairs, ranking reduces to a coin flip. We prove this impossibility, quantify it for four model classes, resolve it via ensemble averaging (DASH), and machine-verify it with 305 Lean 4 theorems. We characterize the complete attribution design space: exactly two families of methods exist -- faithful-complete methods (unstable, with rankings that flip up to 50% of the time) and ensemble methods like DASH (stable, reporting ties for symmetric features) -- and no method lies outside this dichotomy. The impossibility is quantitative: the attribution ratio diverges as 1/(1-rho^2) for gradient boosting, is infinite for Lasso, and converges for random forests. DASH (Diversified Aggregation of SHAP) is provably Pareto-optimal among unbiased aggregations, achieving the Cramer-Rao variance bound with a tight ensemble size formula. In a survey of 77 public datasets, 68% exhibit attribution instability. Switching to conditional SHAP does not escape the impossibility when features have equal causal effects. The framework includes practical diagnostics -- a Z-test workflow and single-model screening tool -- and has direct consequences for fairness auditing: SHAP-based proxy discrimination audits are provably unreliable under collinearity. The design space theorem, diagnostics, and impossibility are mechanically verified in Lean 4 (305 theorems from 16 axioms, 0 sorry) -- to our knowledge, the first formally verified impossibility in explainable AI.
Potential threat/caveat for clean result "Leakage rate is a usable signal for recovering trigger-shaped phrases on Gaperon-1125-1B without knowing the hidden trigger itself (MODERATE confidence)": this item discusses bias.
- score 100arxiv stat.ML (Machine Learning)arxiv:2605.22549unread
A Martingale Kernel Independence Test
Felix Laumann, Zhaolu Liu, Mauricio Barahona · 2026-05-22
arXiv:2605. 22549v1 Announce Type: new Abstract: The Hilbert-Schmidt Independence Criterion (HSIC) and its joint-independence extension $d\mathrm{HSIC}$ are degenerate $V$-statistics whose data-dependent weighted-$\chi^2$ null limits force a permutation calibration that multiplies the per-test cost by the number of permutations, in practice two orders of magnitude.
Read next because A Martingale Kernel Independence Test overlaps with clean result "Language-mismatch LoRA SFT on Qwen2.5-7B leaks the trained completion language into bystander directives the model was never trained on, absent under same-language SFT (LOW confidence)", clean result "The marker is a representational handle, not a behavioural one — sharing it between a villain persona and the assistant transfers no misalignment (HIGH confidence)", experiment "Add C2 control arm (donor sees marker_B without marker_A) to disambiguate paired-marker binding from marker_B leaking alone". Matching terms: under, line, rate, without, test. Source: arxiv stat.ML (Machine Learning).
arXiv:2605.22549v1 Announce Type: new Abstract: The Hilbert-Schmidt Independence Criterion (HSIC) and its joint-independence extension $d\mathrm{HSIC}$ are degenerate $V$-statistics whose data-dependent weighted-$\chi^2$ null limits force a permutation calibration that multiplies the per-test cost by the number of permutations, in practice two orders of magnitude. Adapting the recent martingale MMD construction for two-sample testing to the (joint) independence problem, we introduce two studentised statistics whose null distributions are standard normal regardless of the data law, so that a single normal-quantile lookup replaces the permutation step entirely. The first, $m\mathrm{HSIC}$, is a self-normalised lower-triangular sum of the Hadamard product of two empirically centred Gram matrices. Under independence and bounded-fourth-moment kernels it converges to a standard normal. It is consistent against every fixed alternative, and runs at quadratic cost in the sample size without any sample split, matching the biased HSIC $V$-statistic. Our second statistic, $md\mathrm{HSIC}$, achieves finite-sample consistency with a single half-sample split: the centring is estimated on one half and the lower-triangular self-normalised martingale is run on the other, shrinking the conditional-mean residual to a quantity that is exponentially small in $d$, so the statistic is asymptotically standard normal at every fixed number of jointly tested variables, with a per-test cost that grows only linearly in $d$. On synthetic data with per-variable input dimension from $1$ to $500$ and between $2$ and $10$ jointly tested variables, both statistics match the empirical type-I error rate and test power of permutation-calibrated baselines while running $25$ to $60\times$ faster.
Potential threat/caveat for clean result "Language-mismatch LoRA SFT on Qwen2.5-7B leaks the trained completion language into bystander directives the model was never trained on, absent under same-language SFT (LOW confidence)": this item discusses bias.
- score 100arxiv stat.ML (Machine Learning)arxiv:2605.22438unread
Do Not Trust The Auctioneer: Learning to Bid in Feedback-Manipulated Auctions
Luigi Foscari, Matilde Tullii, Vianney Perchet · 2026-05-22
arXiv:2605. 22438v1 Announce Type: new Abstract: Shilling is the use of artificial bids to make competition appear stronger and push prices upward.
Read next because Do Not Trust The Auctioneer: Learning to Bid in Feedback-Manipulated Auctions overlaps with clean result "LoRA persona trained on <A> alone emits <B> at 23.5% when a co-trained partner learns <A>...<B>, vs 0% control on Qwen2.5-7B-Instruct (MODERATE confidence)", clean result "Leakage rate is a usable signal for recovering trigger-shaped phrases on Gaperon-1125-1B without knowing the hidden trigger itself (MODERATE confidence)", clean result "Language-mismatch LoRA SFT on Qwen2.5-7B leaks the trained completion language into bystander directives the model was never trained on, absent under same-language SFT (LOW confidence)". Matching terms: strong, rate, without, factor, suffix. Source: arxiv stat.ML (Machine Learning).
arXiv:2605.22438v1 Announce Type: new Abstract: Shilling is the use of artificial bids to make competition appear stronger and push prices upward. We study repeated first-price auctions in which shilling affects feedback but not allocation: the learner wins or loses against the real competing bid, but after a loss observes the maximum of the real bid and an independent shill bid. Thus the manipulation changes what the learner observes and hence how it learns to bid, without changing the outcome of the current auction. We analyze regret with respect to the best bid benchmark, assuming that the shill-bid distribution is known. Even then, shilling can mask the real bid, while useful side information appears only through intermittent low-shill events. Our algorithm combines a robust interval-elimination branch, which ignores the shilled report and achieves the dynamic-pricing rate $\tilde{\mathcal{O}}(T^{2/3})$, with an optimistic branch that debiases losing-side reports and exploits the resulting suffix information when it is reliable and achieves the first-price auctions rate $\tilde{\mathcal{O}}(\sqrt{T})$. A validation and racing procedure lets the algorithm use these optimistic updates without knowing the right scale or feedback geometry in advance. We complement the upper bounds with a matching lower bound, up to logarithmic factors, in the single-active-region case. Overall, the results show that even feedback-only shilling can sharply alter the statistical difficulty of repeated bidding.
Potential threat/caveat for clean result "LoRA persona trained on <A> alone emits <B> at 23.5% when a co-trained partner learns <A>...<B>, vs 0% control on Qwen2.5-7B-Instruct (MODERATE confidence)": this item discusses bias, benchmark.
- score 100arxiv stat.ML (Machine Learning)arxiv:2605.21736unread
Support-aware offline policy selection for advertising marketplaces
Prashant Shekhar, Caroline Howard · 2026-05-22
arXiv:2605. 21736v1 Announce Type: new Abstract: Logged advertising auctions make offline reserve-price evaluation attractive but risky.
Read next because Support-aware offline policy selection for advertising marketplaces overlaps with clean result "LoRA persona trained on <A> alone emits <B> at 23.5% when a co-trained partner learns <A>...<B>, vs 0% control on Qwen2.5-7B-Instruct (MODERATE confidence)", clean result "Leakage rate is a usable signal for recovering trigger-shaped phrases on Gaperon-1125-1B without knowing the hidden trigger itself (MODERATE confidence)", clean result "Language-mismatch LoRA SFT on Qwen2.5-7B leaks the trained completion language into bystander directives the model was never trained on, absent under same-language SFT (LOW confidence)". Matching terms: strong, rect, under, eval, line, control, alone, candidates. Source: arxiv stat.ML (Machine Learning).
arXiv:2605.21736v1 Announce Type: new Abstract: Logged advertising auctions make offline reserve-price evaluation attractive but risky. Replay tables can identify policies with large apparent yield gains, yet they can also hide weak threshold support, multiple-comparison effects, subgroup harm, and bidder-response uncertainty. Existing replay and off-policy evaluation methods estimate or rank policy values, but they do not directly answer the operational question of whether the available evidence is strong enough to justify validation. This paper develops a support-aware offline decision framework for reserve-policy selection. Rather than outputting a single point-estimate winner, the framework converts logged evidence into a conservative decision object consisting of certified policies, statistically dominated alternatives, and unresolved candidates requiring further validation. The main theoretical result gives a unified finite-catalog guarantee showing that, under simultaneous uncertainty control and conservative support gates, the framework preserves the best gate-passing policy while eliminating only policies with certified regret. Supporting results characterize support-localized replay generalization, establish information-theoretic threshold-resolution limits, and quantify when heterogeneous bidder response can overturn localized replay rankings. Experiments on iPinYou real-time-bidding logs show that the leading reserve rule achieves a 47.66% replay lift in season two, a 40.71% simultaneous lower-bound lift, and a 43.87% frozen out-of-time replay lift in season three. The framework reduces a 19-policy catalog to a two-policy validation shortlist while certifying non-harm across 44 advertiser, exchange, and region segments. The results support the central claim that offline reserve-policy evaluation should produce certified validation decisions rather than point-estimate rankings alone.
Potential threat/caveat for clean result "LoRA persona trained on <A> alone emits <B> at 23.5% when a co-trained partner learns <A>...<B>, vs 0% control on Qwen2.5-7B-Instruct (MODERATE confidence)": this item discusses evaluation.
- score 100arxiv stat.ML (Machine Learning)arxiv:2605.21557unread
Scalable On-Policy Reinforcement Learning via Adaptive Batch Scaling
Jongchan Park · 2026-05-22
arXiv:2605. 21557v1 Announce Type: new Abstract: Conventional wisdom holds that large-batch training is fundamentally incompatible with Reinforcement Learning (RL) - beyond a modest threshold, increasing batch sizes typically yields diminishing returns or performance degradation due to the inherent non-stationarity of the data distribution.
Read next because Scalable On-Policy Reinforcement Learning via Adaptive Batch Scaling overlaps with clean result "Only continuous soft prefixes hit both EM axes at once on Qwen-2.5-7B-Instruct: discrete prompt searches split between the alignment objective and the distributional objective, and both discretizations of the soft prefix collapse (MODERATE confidence)", clean result "The marker is a representational handle, not a behavioural one — sharing it between a villain persona and the assistant transfers no misalignment (HIGH confidence)", experiment "Add C2 control arm (donor sees marker_B without marker_A) to disambiguate paired-marker binding from marker_B leaking alone". Matching terms: eval, rate, control, on-policy, stage. Source: arxiv stat.ML (Machine Learning).
arXiv:2605.21557v1 Announce Type: new Abstract: Conventional wisdom holds that large-batch training is fundamentally incompatible with Reinforcement Learning (RL) - beyond a modest threshold, increasing batch sizes typically yields diminishing returns or performance degradation due to the inherent non-stationarity of the data distribution. We challenge this view by observing that non-stationarity is not a fixed property of RL, but evolves throughout training: early stages exhibit rapid behavioral shifts that demand small batches for plasticity, whereas late stages approach a quasi-stationary regime where large batches enable precise convergence. Motivated by this observation, we propose Adaptive Batch Scaling (ABS), that dynamically adjusts the effective batch size according to the stability of the learning policy. Central to ABS is Behavioral Divergence, a novel metric that quantifies policy non-stationarity by measuring action-level shifts between consecutive updates, which we use to scale batch size inversely to policy volatility. Integrated with the Parallelised Q-Network (PQN) algorithm and evaluated on the ALE benchmark, ABS seamlessly reconciles early-stage plasticity with late-stage stable convergence. Strikingly, contrary to conventional wisdom, our results reveal that the combination of larger networks and larger batch sizes achieves the best performance - a scaling behavior previously thought to be unattainable in RL, now unlocked through adaptive batch control.
Potential threat/caveat for clean result "Only continuous soft prefixes hit both EM axes at once on Qwen-2.5-7B-Instruct: discrete prompt searches split between the alignment objective and the distributional objective, and both discretizations of the soft prefix collapse (MODERATE confidence)": this item discusses benchmark.
- score 100arxiv stat.ML (Machine Learning)arxiv:2605.21548unread
Local Covariate Selection for Average Causal Effect Estimation without Pretreatment and Causal Sufficiency Assumptions
Zeyu Liu, Zheng Li, Feng Xie, Yan Zeng, Hao Zhang, Kun Zhang · 2026-05-22
arXiv:2605. 21548v1 Announce Type: new Abstract: We study the problem of selecting covariates for unbiased estimation of the total causal effect.
Read next because Local Covariate Selection for Average Causal Effect Estimation without Pretreatment and Causal Sufficiency Assumptions overlaps with clean result "LoRA persona trained on <A> alone emits <B> at 23.5% when a co-trained partner learns <A>...<B>, vs 0% control on Qwen2.5-7B-Instruct (MODERATE confidence)", clean result "Leakage rate is a usable signal for recovering trigger-shaped phrases on Gaperon-1125-1B without knowing the hidden trigger itself (MODERATE confidence)", clean result "Language-mismatch LoRA SFT on Qwen2.5-7B leaks the trained completion language into bystander directives the model was never trained on, absent under same-language SFT (LOW confidence)". Matching terms: strong, under, rate, without, never. Source: arxiv stat.ML (Machine Learning).
arXiv:2605.21548v1 Announce Type: new Abstract: We study the problem of selecting covariates for unbiased estimation of the total causal effect.Existing approaches typically rely on global causal structure learning over all variables, or on strong assumptions such as causal sufficiency - where observed variables share no latent confounders - or the pretreatment assumption, which limits covariates to those unaffected by the treatment or outcome. These requirements are often unrealistic in practice, and global learning becomes computationally prohibitive in high-dimensional settings.To address these challenges, we propose a novel local learning method for covariate selection in nonparametric causal effect estimation that avoids both the pretreatment and causal sufficiency assumptions. We first characterize a local boundary that contains at least one valid adjustment set whenever one exists for identifying the causal effect, and then develop local identification procedures to efficiently search within this boundary.We prove that the proposed method is sound and complete. Experiments on multiple synthetic datasets and two real-world datasets show that our approach achieves accurate causal effect estimation while substantially improving computational efficiency.
Potential threat/caveat for clean result "LoRA persona trained on <A> alone emits <B> at 23.5% when a co-trained partner learns <A>...<B>, vs 0% control on Qwen2.5-7B-Instruct (MODERATE confidence)": this item discusses bias, confound.
- score 100arxiv stat.ML (Machine Learning)arxiv:2605.21534unread
Adaptive RBF-KAN: A Comparative Evaluation of Dynamic Shape Parameters in Kolmogorov-Arnold Networks
Roberto Cavoretto, Alessandra De Rossi, Adeeba Haider, Amir Noorizadegan · 2026-05-22
arXiv:2605. 21534v1 Announce Type: new Abstract: Kolmogorov-Arnold Networks (KANs) approximate multivariate functions using learnable univariate edge functions, typically parameterized by B-spline bases.
Read next because Adaptive RBF-KAN: A Comparative Evaluation of Dynamic Shape Parameters in Kolmogorov-Arnold Networks overlaps with clean result "Only continuous soft prefixes hit both EM axes at once on Qwen-2.5-7B-Instruct: discrete prompt searches split between the alignment objective and the distributional objective, and both discretizations of the soft prefix collapse (MODERATE confidence)", clean result "The marker is a representational handle, not a behavioural one — sharing it between a villain persona and the assistant transfers no misalignment (HIGH confidence)", experiment "Implement Chen et al. persona-vector extraction recipe and compare to project's centroid-difference recipe". Matching terms: eval, line, rate, implement, model. Source: arxiv stat.ML (Machine Learning).
arXiv:2605.21534v1 Announce Type: new Abstract: Kolmogorov-Arnold Networks (KANs) approximate multivariate functions using learnable univariate edge functions, typically parameterized by B-spline bases. Although effective, spline-based implementations can be computationally expensive. A modified version of KANs, called FastKAN, improves efficiency by replacing splines with Gaussian radial basis functions (RBFs), but it relies on a fixed kernel and shape parameter. In this work, we extend the RBF-based KAN framework by introducing a broader family of radial basis kernels and by initializing the kernel shape parameter using leave-one-out cross-validation (LOOCV). To the best of our knowledge, this is the first study that integrates LOOCV-based kernel scale estimation with deep KAN training. We also introduce Mat\'ern and Wendland kernels into the KAN framework for the first time, enabling more flexible basis representations beyond the Gaussian kernel used in FastKAN. The LOOCV estimate provides a data-driven initialization of the kernel scale, which is subsequently refined during network training. The proposed adaptive RBF-KAN is evaluated on several two-dimensional benchmark functions. The results highlight the importance of kernel selection and adaptive shape parameters, with different kernels showing advantages for smooth functions, discontinuities, and oscillatory patterns. Overall, combining LOOCV-based initialization with adaptive kernel learning provides a practical strategy for improving RBF-based KAN models.
Potential threat/caveat for clean result "Only continuous soft prefixes hit both EM axes at once on Qwen-2.5-7B-Instruct: discrete prompt searches split between the alignment objective and the distributional objective, and both discretizations of the soft prefix collapse (MODERATE confidence)": this item discusses evaluation, benchmark.
- score 64arxiv stat.ML (Machine Learning)arxiv:2605.21860unread
Robust Statistical Estimators with Bounded Empirical Sensitivity
Valentio Iverson, Gautam Kamath, Argyris Mouzakis, Adam Smith · 2026-05-22
arXiv:2605. 21860v1 Announce Type: cross Abstract: We introduce a new measure of robustness for statistical estimators, which we call \emph{empirical sensitivity}.
Read next because Robust Statistical Estimators with Bounded Empirical Sensitivity overlaps with experiment "Factor screen for marker implantation + leakage (2^5: system-prompt length, answer-format length, persona-presence, on-policy, marker-only-loss)". Matching terms: factor. Source: arxiv stat.ML (Machine Learning).
arXiv:2605.21860v1 Announce Type: cross Abstract: We introduce a new measure of robustness for statistical estimators, which we call \emph{empirical sensitivity}. An estimator $\hat \theta$ has bounded empirical sensitivity if, with high probability over a dataset $X = (X_1, \dots, X_n) \sim \mathcal{D}^{\otimes n}$, for any dataset $Y$ obtained by modifying at most $\eta n$ points in $X$, we have that $\hat \theta(Y)$ is close to $\hat \theta(X)$. We study bounds on this quantity for the prototypical problem of Gaussian mean estimation. We prove new lower bounds, showing that for any estimator $\hat \mu$ which achieves an optimal $\ell_2$-error bound of $O\left(\sqrt{d/n}\right)$, the empirical sensitivity is at least $\Omega\left(\eta + \sqrt{\eta d/n}\right)$. The two terms arise due to obstructions on the mean and variance (via an Efron-Stein argument) of such an estimator. We show that this bound is tight up to logarithmic factors, by employing recent results for robust empirical mean estimation.
Potential threat/caveat for experiment "Factor screen for marker implantation + leakage (2^5: system-prompt length, answer-format length, persona-presence, on-policy, marker-only-loss)": this item discusses robustness.
Methods
- score 38M7 QA inline RSS threat sourceunread
Artifact verification caveats for Sagan clean results
M7 QA · No release date
This paper studies failure modes and caveats when Sagan creates a clean result only after verifying an artifact row. It proposes benchmark checks for artifact verification, clean-result review comments, and negative controls.
My clean results (persona marker leakage, bystander language spill, backdoor token-binding, etc.) all passed through artifact verification before receiving confidence labels; this document directly concerns whether that verification step is trustworthy and what systematic failure modes I should watch for in my own pipeline.
This paper studies failure modes and caveats when Sagan creates a clean result only after verifying an artifact row. It proposes benchmark checks for artifact verification, clean-result review comments, and negative controls.
Potential threat/caveat for experiment "Add C2 control arm (donor sees marker_B without marker_A) to disambiguate paired-marker binding from marker_B leaking alone": this item discusses failure, caveat, caveats, negative, benchmark.