Daily reading queue

31 items for 2026-05-16 across 3 categories.

Active sources: 7. Sources represented in this queue: 2. The cron runs daily at 06:00 server time; arxiv RSS is often empty on weekends.

Linked to your results

score 100arxiv cs.AI (Artificial Intelligence)arxiv:2605.14212unread
MetaAgent-X : Breaking the Ceiling of Automatic Multi-Agent Systems via End-to-End Reinforcement Learning
Yaolun Zhang, Yujie Zhao, Nan Wang, Yiran Wu, Jiayu Chang, Yizhao Chen, Qingyun Wu, Jishen Zhao, Huazheng Wang · 2026-05-16
arXiv:2605. 14212v1 Announce Type: new Abstract: Automatic multi-agent systems aim to instantiate agent workflows without relying on manually designed or fixed orchestration.
Read next because MetaAgent-X : Breaking the Ceiling of Automatic Multi-Agent Systems via End-to-End Reinforcement Learning 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: line, without, stage, test, model. Source: arxiv cs.AI (Artificial Intelligence).
arXiv:2605.14212v1 Announce Type: new Abstract: Automatic multi-agent systems aim to instantiate agent workflows without relying on manually designed or fixed orchestration. However, existing automatic MAS approaches remain only partially adaptive: they either perform training-free test-time search or optimize the meta-level designer while keeping downstream execution agents frozen, which creating a frozen-executor ceiling and leaving the end-to-end training of self-designing and self-executing agentic models unexplored. To address this, we introduce MetaAgent-X, an end-to-end reinforcement learning framework that jointly optimizes automatic MAS design and execution. MetaAgent-X enables script-based MAS generation, execution rollout collection, and credit assignment for both designer and executor trajectories. To support stable and scalable optimization, we propose Executor Designer Hierarchical Rollout and Stagewise Co-evolution to improve training stability and expose the dynamics of designer-executor co-evolution. MetaAgent-X consistently outperforms existing automatic MAS baselines, achieving up to 21.7% gains. Comprehensive ablations show that both designer and executor improve throughout training, and that effective automatic MAS learning follows a stagewise co-evolution process. These results establish end-to-end trainable automatic MAS as a practical paradigm for building self-designing and self-executing agentic models.
source ↗open
score 100arxiv cs.AI (Artificial Intelligence)arxiv:2605.14205unread
SimPersona: Learning Discrete Buyer Personas from Raw Clickstreams for Grounded E-Commerce Agents
Zahra Zanjani Foumani, Alberto Castelo, Shuang Xie, Ted Chaiwachirasak, Han Li, Lingyun Wang · 2026-05-16
arXiv:2605. 14205v1 Announce Type: new Abstract: LLM-based web agents can navigate live storefronts, yet they often collapse to a single "average buyer" policy, failing to capture the heterogeneous and distributional nature of real buyer populations.
Read next because SimPersona: Learning Discrete Buyer Personas from Raw Clickstreams for Grounded E-Commerce Agents overlaps with clean result "LoRA persona trained on <A> alone emits at 23.5% when a co-trained partner learns <A>..., 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, alignment, distributional, eval, source, token. Source: arxiv cs.AI (Artificial Intelligence).
arXiv:2605.14205v1 Announce Type: new Abstract: LLM-based web agents can navigate live storefronts, yet they often collapse to a single "average buyer" policy, failing to capture the heterogeneous and distributional nature of real buyer populations. Existing personalization methods rely on hand-crafted prompt-based personas that are brittle, difficult to scale, context-inefficient, and unable to faithfully represent population-level behavior. We introduce SimPersona, a novel framework that learns discrete buyer types from historical traffic and exposes them to LLM-based web agents as compact persona tokens. Given raw clickstreams, a behavior-aware VQ-VAE induces a discrete buyer-type space that captures the statistical structure of real buyer behavior and merchant-specific buyer population distributions. To provide behavior-specific guidance to LLM-based web agents, SimPersona maps each learned buyer type to a dedicated persona token in the LLM agent vocabulary and fine-tunes the agent with these tokens on real browsing traces. At inference, each synthetic buyer is assigned to a learned buyer type with a single encoder forward pass, requiring no retraining or store-specific prompt engineering. For population-level simulation, SimPersona samples buyer types from each merchant's empirical distribution over the learned VQ-VAE codebook and instantiates agents with the corresponding persona tokens, preserving merchant-specific buyer population distributions. Evaluated on $8.37$M buyers across $42$ held-out live storefronts, SimPersona achieves $78\%$ conversion-rate alignment with real buyers, exhibits interpretable behavioral variation across buyer types, and outperforms a baseline with $8\times$ more parameters on goal-oriented shopping tasks. We further release an open-source data pipeline that converts raw e-commerce event logs into buyer representations and agent-training traces.
source ↗open
score 100arxiv cs.AI (Artificial Intelligence)arxiv:2605.14141unread
Distribution-Aware Algorithm Design with LLM Agents
Saharsh Koganti, Priyadarsi Mishra, Pierfrancesco Beneventano, Tomer Galanti · 2026-05-16
arXiv:2605. 14141v1 Announce Type: new Abstract: We study learning when the learned object is executable solver code rather than a predictor.
Read next because Distribution-Aware Algorithm Design with LLM Agents overlaps with clean result "LoRA persona trained on <A> alone emits at 23.5% when a co-trained partner learns <A>..., 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, correct, eval, rate, test. Source: arxiv cs.AI (Artificial Intelligence).
arXiv:2605.14141v1 Announce Type: new Abstract: We study learning when the learned object is executable solver code rather than a predictor. In this setting, correctness is not enough: two solvers may both return valid solutions on the deployment distribution while differing substantially in runtime. Given samples from an unknown task distribution, the learner returns code evaluated on fresh instances by both solution quality and execution time. Our central abstraction is a \emph{solver hint}: reusable structure inferred from samples and compiled into specialized solver code. We prove that the empirically fastest sample-consistent solver from a fixed library generalizes in both correctness and runtime, and that statistically identifiable hints can be recovered and compiled from polynomially many samples. Empirically, we instantiate the framework with LLM code agents on $21$ structured combinatorial-optimization target distributions across seven problem classes. The synthesized solvers reach mean normalized quality $0.971$, improve by $+0.224$ over the average heuristic pool and by $+0.098$ over the highest-quality heuristic, and are $336.9\times$, $342.8\times$, and $16.1\times$ faster than the quality-best heuristic, Gurobi, and the selected time-limited exact backend, respectively. On released PACE 2025 Dominating Set private instances, the synthesized solver is valid on all $100$ graphs and runs about two orders of magnitude faster than top competition solvers, with a moderate quality gap. Inspection shows that many gains come from changing the computational scale: replacing ambient exponential search or general-purpose optimization with compiled distribution-specific computation.
source ↗open
score 100arxiv cs.AI (Artificial Intelligence)arxiv:2605.14111unread
Modeling Bounded Rationality in Drug Shortage Pharmacists Using Attention-Guided Dynamic Decomposition
Yaniv Eliyahu Amiri, Noah Chicoine, Jacqueline Griffin, Stacy Marsella · 2026-05-16
arXiv:2605. 14111v1 Announce Type: new Abstract: Hospital pharmacists make high-stakes decisions to mitigate drug shortages under uncertainty, time pressure, and patient risk.
Read next because Modeling Bounded Rationality in Drug Shortage Pharmacists Using Attention-Guided Dynamic Decomposition 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, rate, without, position, model. Source: arxiv cs.AI (Artificial Intelligence).
arXiv:2605.14111v1 Announce Type: new Abstract: Hospital pharmacists make high-stakes decisions to mitigate drug shortages under uncertainty, time pressure, and patient risk. Interviews revealed that pharmacists focus attention on a small subset of drugs, limiting cognitive effort to the most urgent cases. Motivated by these findings, we formalize a bounded-rational, attention-guided decision framework that dynamically decomposes drugs into a subset for high-cost reasoning and a complementary subset for low-cost monitoring. We develop two agents: an Expert Agent that applies attention weights derived from pharmacist interviews, and a Learner Agent that adapts attention allocation over time through experience. Across simulated scenarios spanning short to long horizons, we show that attention-guided planning supports stable decision-making without complete state reasoning. These results suggest that a primary decision is not what action to take, but where to allocate cognitive effort, and that attention-guided, satisficing strategies can reduce problem complexity while maintaining stable performance.
source ↗open
score 100arxiv cs.AI (Artificial Intelligence)arxiv:2605.14089unread
SkillFlow: Flow-Driven Recursive Skill Evolution for Agentic Orchestration
Mingda Zhang, Tiesunlong Shen, Haoran Luo, Wenjin Liu, Zikai Xiao, Erik Cambria, Xiaoying Tang · 2026-05-16
arXiv:2605. 14089v1 Announce Type: new Abstract: In recent years, a variety of powerful LLM-based agentic systems have been applied to automate complex tasks through task orchestration.
Read next because SkillFlow: Flow-Driven Recursive Skill Evolution for Agentic Orchestration overlaps with clean result "LoRA persona trained on <A> alone emits at 23.5% when a co-trained partner learns <A>..., 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, under, line, rate, capability. Source: arxiv cs.AI (Artificial Intelligence).
arXiv:2605.14089v1 Announce Type: new Abstract: In recent years, a variety of powerful LLM-based agentic systems have been applied to automate complex tasks through task orchestration. However, existing orchestration methods still face key challenges, including strategy collapse under reward maximization, high gradient variance with opaque credit assignment, and unguided skill evolution whose decisions are typically made by directly prompting an LLM to judge rather than derived from principled training signals. To address these challenges, we propose SkillFlow, a flow-based framework that takes a trainable Supervisor as the agent and a structured environment with dynamic skill library and frozen executor, automating task orchestration through multi-turn interaction. SkillFlow employs Tempered Trajectory Balance (TTB), a regression-based flow-matching loss that samples trajectories proportional to reward, preserving diverse orchestration strategies rather than collapsing to a single mode. The same flow objective yields a jointly learned backward policy that provides transparent per-step credit assignment at zero additional inference cost. Building on these flow diagnostics, a recursive skill evolution mechanism determines when to evolve, what skills to create or prune, and where decision gaps lie -- closing the loop from training signal to autonomous capability growth. Experimental results on 14 datasets show that SkillFlow significantly outperforms baselines across question answering, mathematical reasoning, code generation, and real-world interactive decision making tasks. Our code is available at https://anonymous.4open.science/r/SkillFlow-E850.
source ↗open
score 100arxiv cs.AI (Artificial Intelligence)arxiv:2605.14054unread
Bad Seeing or Bad Thinking? Rewarding Perception for Vision-Language Reasoning
Haozhe Wang, Qixin Xu, Changpeng Wang, Taofeng Xue, Chong Peng, Wenhu Chen, Fangzhen Lin · 2026-05-16
arXiv:2605. 14054v1 Announce Type: new Abstract: Achieving robust perception-reasoning synergy is a central goal for advanced Vision-Language Models (VLMs).
Read next because Bad Seeing or Bad Thinking? Rewarding Perception for Vision-Language Reasoning overlaps with clean result "LoRA persona trained on <A> alone emits at 23.5% when a co-trained partner learns <A>..., 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, source, rate, sees, does, language, model. Source: arxiv cs.AI (Artificial Intelligence).
arXiv:2605.14054v1 Announce Type: new Abstract: Achieving robust perception-reasoning synergy is a central goal for advanced Vision-Language Models (VLMs). Recent advancements have pursued this goal via architectural designs or agentic workflows. However, these approaches are often limited by static textual reasoning or complicated by the significant compute and engineering burden of external agentic complexity. Worse, this heavy investment does not yield proportional gains, often witnessing a "seesaw effect" on perception and reasoning. This motivates a fundamental rethinking of the true bottleneck. In this paper, we argue that the root cause of this trade-off is an ambiguity in modality credit assignment: when a VLM fails, is it due to flawed perception ("bad seeing") or flawed logic ("bad thinking")? To resolve this, we introduce a reinforcement learning framework that improves perception-reasoning synergy by reliably rewarding the perception fidelity. We explicitly decompose the generation process into interleaved perception and reasoning steps. This decoupling enables targeted supervision on perception. Crucially, we introduce Perception Verification (PV), leveraging a "blindfolded reasoning" proxy to reward perceptual fidelity independently of reasoning outcomes. Furthermore, to scale training across free-form VL tasks, we propose Structured Verbal Verification, which replaces high-variance LLM judging with structured algorithmic execution. These techniques are integrated into a Modality-Aware Credit Assignment (MoCA) mechanism, which routes rewards to the specific source of error -- either bad seeing or bad thinking -- enabling a single VLM to achieve simultaneous performance gains across a wide task spectrum.
source ↗open
score 100arxiv cs.AI (Artificial Intelligence)arxiv:2605.14049unread
Bridging Legal Interpretation and Formal Logic: Faithfulness, Assumption, and the Future of AI Legal Reasoning
Olivia Peiyu Wang, Leilani H. Gilpin · 2026-05-16
arXiv:2605. 14049v1 Announce Type: new Abstract: The growing adoption of large language models in legal practice brings both significant promise and serious risk.
Read next because Bridging Legal Interpretation and Formal Logic: Faithfulness, Assumption, and the Future of AI Legal Reasoning overlaps with clean result "LoRA persona trained on <A> alone emits at 23.5% when a co-trained partner learns <A>..., 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, source, without, language, model. Source: arxiv cs.AI (Artificial Intelligence).
arXiv:2605.14049v1 Announce Type: new Abstract: The growing adoption of large language models in legal practice brings both significant promise and serious risk. Legal professionals stand to benefit from AI that can reason over contracts, draft documents, and analyze sources at scale, yet the high-stakes nature of legal work demands a level of rigor that current AI systems do not provide. The central problem is not simply that LLMs hallucinate facts and references; it is that they systematically draw inferences that go beyond what the source text actually supports, presenting assumption-laden conclusions as if they were logically grounded. This proposal presents a neuro-symbolic approach to legal AI that combines the expressive power of large language models with the rigor of formal verification, aiming to make AI-assisted legal reasoning both capable and trustworthy, thus reducing the burden of manual verification without sacrificing the accountability that legal practice demands.
source ↗open
score 100arxiv cs.AI (Artificial Intelligence)arxiv:2605.14036unread
Enhanced and Efficient Reasoning in Large Learning Models
Leslie G. Valiant · 2026-05-16
arXiv:2605. 14036v1 Announce Type: new Abstract: In current Large Language Models we can trust the production of smoothly flowing prose on the basis of the principles of machine learning.
Read next because Enhanced and Efficient Reasoning in Large Learning Models overlaps with clean result "LoRA persona trained on <A> alone emits at 23.5% when a co-trained partner learns <A>..., 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, soft, line, chain, stage, language. Source: arxiv cs.AI (Artificial Intelligence).
arXiv:2605.14036v1 Announce Type: new Abstract: In current Large Language Models we can trust the production of smoothly flowing prose on the basis of the principles of machine learning. However, there is no comparably principled basis to justify trust in the content of the text produced. It appears to be conventional wisdom that addressing this issue by adding more principled reasoning is not computationally affordable. Here we propose a principled method of reasoning that is efficient enough to be practical for large language models. Further, the method allows the retention of much of the currently used software and hardware base. Our method for improving the functioning of large language models consists of a first stage of preprocessing that recodes the data to a Unary Relational Integracode that is more explicit about the relationships among the objects described in the text, followed as a second stage by a standard but possibly streamlined machine learning process that then also learns to predict these relationships. The method may be viewed as realizing a world model and applying beyond natural language, to vision and actions, for example, where the multiple properties of an object referred to in an input are brought together explicitly, rather than remaining distributed in the various references to it in the input. We articulate its advantages in terms of Robust Logic, a system for performing principled chaining on learned, and hence uncertain, information. We show that this recoding has the surprising and fortuitous property that, while succinct, it makes the task of learning a core subset of relational rules that hold in the world described in the training data polynomial time learnable in a defined sense, the polynomial depending on the complexity of the rule. This gives support for sound reasoning within each single call of the learned classifier as well as between multiple calls.
source ↗open
score 100arxiv cs.AI (Artificial Intelligence)arxiv:2605.14004unread
Conditional Attribute Estimation with Autoregressive Sequence Models
Erica Stutz, Giacomo Marino, Daniella Meeker, Qiao Liu, Andrew J. Loza · 2026-05-16
arXiv:2605. 14004v1 Announce Type: new Abstract: Generative models are often trained with a next-token prediction objective, yet many downstream applications require the ability to estimate or control sequence-level properties.
Read next because Conditional Attribute Estimation with Autoregressive 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 "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: under, token, rate, control, without, trained, language, model. Source: arxiv cs.AI (Artificial Intelligence).
arXiv:2605.14004v1 Announce Type: new Abstract: Generative models are often trained with a next-token prediction objective, yet many downstream applications require the ability to estimate or control sequence-level properties. Next-token prediction can lead to overfitting of local patterns during training, underfitting of global structure, and requires significant downstream modifications or expensive sampling to guide or predict the global attributes of generated samples at inference time. Here, we introduce Conditional Attribute Transformers, a novel method for jointly estimating the next-token probability and the value of an attribute conditional on each potential next token selection. This framework enables three critical capabilities within a single forward pass, without modification of the input sequence: (1) per-token credit assignment across an entire sequence, by identifying how each token in a sequence is associated with an attribute's value; (2) counterfactual analysis, by quantifying attribute differences conditional on alternative next token choices; (3) steerable generation, by decoding sequences based on a combination of next-token and attribute likelihoods. Our approach achieves state of the art performance on sparse reward tasks, improves next-token prediction at sufficient model sizes, estimates attribute probabilities orders of magnitude faster than sampling, and can guide decoding of autoregressive sequence models on a range of language tasks.
source ↗open
score 100arxiv cs.AI (Artificial Intelligence)arxiv:2605.13880unread
PREPING: Building Agent Memory without Tasks
Yumin Choi, Sangwoo Park, Minki Kang, Jinheon Baek, Sung Ju Hwang · 2026-05-16
arXiv:2605. 13880v1 Announce Type: new Abstract: Agent memory is typically constructed either offline from curated demonstrations or online from post-deployment interactions.
Read next because PREPING: Building Agent Memory without Tasks overlaps with clean result "LoRA persona trained on <A> alone emits at 23.5% when a co-trained partner learns <A>..., 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, control, without, alone, does. Source: arxiv cs.AI (Artificial Intelligence).
arXiv:2605.13880v1 Announce Type: new Abstract: Agent memory is typically constructed either offline from curated demonstrations or online from post-deployment interactions. However, regardless of how it is built, an agent faces a cold-start gap when first introduced to a new environment without any task-specific experience available. In this paper, we study pre-task memory construction: whether an agent can build procedural memory before observing any target-environment tasks, using only self-generated synthetic practice. Yet, synthetic interaction alone is insufficient, as without controlling what to practice and what to store, synthetic tasks become redundant, infeasible, and ultimately uninformative, and memory further degrades quickly due to unfiltered trajectories. To overcome this, we present Preping, a proposer-guided memory construction framework. At its core is proposer memory, a structured control state that shapes future practice. A Proposer generates synthetic tasks conditioned on this state, a Solver executes them, and a Validator determines which trajectories are eligible for memory insertion while also providing feedback to guide future proposals. Experiments on AppWorld, BFCL v3, and MCP-Universe show that Preping substantially improves over a no-memory baseline and achieves performance competitive with strong playbook-based methods built from offline or online experience, with deployment cost $2.99\times$ lower on AppWorld and $2.23\times$ lower on BFCL v3 than online memory construction. Further analyses reveal that the main benefit does not come from synthetic volume alone, but from proposer-side control over feasibility, redundancy, and coverage, combined with selective memory updates.
source ↗open

Threats and caveats

score 100arxiv cs.AI (Artificial Intelligence)arxiv:2605.14215unread
GenCircuit-RL: Reinforcement Learning from Hierarchical Verification for Genetic Circuit Design
Noah Flynn · 2026-05-16
arXiv:2605. 14215v1 Announce Type: new Abstract: Genetic circuit design remains a laborious, expert-driven process despite decades of progress in synthetic biology.
Read next because GenCircuit-RL: Reinforcement Learning from Hierarchical Verification for Genetic Circuit Design overlaps with clean result "LoRA persona trained on <A> alone emits at 23.5% when a co-trained partner learns <A>..., 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, rect, correct, eval, rate, stage, language. Source: arxiv cs.AI (Artificial Intelligence).
arXiv:2605.14215v1 Announce Type: new Abstract: Genetic circuit design remains a laborious, expert-driven process despite decades of progress in synthetic biology. We study this problem through code generation: models produce Python code in pysbol3 to construct genetic circuits in the Synthetic Biology Open Language (SBOL), a formal representation that supports automated verification. We introduce GenCircuit-RL, a reinforcement learning framework built around hierarchical verification rewards that decompose correctness into five levels, from code execution to task-specific topological checks, and a four-stage curriculum that shifts optimization pressure from code generation to functional reasoning. We also introduce SynBio-Reason, a benchmark of 4,753 circuits spanning six canonical circuit types and nine tasks from code repair to de novo design, with held-out biological parts for out-of-distribution evaluation. Hierarchical verification improves task success on functional reasoning tasks by 14 to 16 percentage points over binary rewards, and curriculum learning is required for strong design performance. The resulting models generate topologically correct circuits, generalize to novel biological parts, and rediscover canonical designs from the synthetic biology literature.
Potential threat/caveat for clean result "LoRA persona trained on <A> alone emits at 23.5% when a co-trained partner learns <A>..., vs 0% control on Qwen2.5-7B-Instruct (MODERATE confidence)": this item discusses evaluation, benchmark.
source ↗open
score 100arxiv cs.AI (Artificial Intelligence)arxiv:2605.14211unread
ASH: Agents that Self-Hone via Embodied Learning
Benjamin Schneider, Xavier Schneider, Victor Zhong, Sun Sun · 2026-05-16
arXiv:2605. 14211v1 Announce Type: new Abstract: Long-horizon embodied tasks remain a fundamental challenge in AI, as current methods rely on hand-engineered rewards or action-labeled demonstrations, neither of which scales.
Read next because ASH: Agents that Self-Hone via Embodied Learning overlaps with clean result "LoRA persona trained on <A> alone emits at 23.5% when a co-trained partner learns <A>..., 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, eval, line, rate, recipe, without, model. Source: arxiv cs.AI (Artificial Intelligence).
arXiv:2605.14211v1 Announce Type: new Abstract: Long-horizon embodied tasks remain a fundamental challenge in AI, as current methods rely on hand-engineered rewards or action-labeled demonstrations, neither of which scales. We introduce ASH, an agentic system that learns an embodied policy from unlabeled, noisy internet video, without reward shaping or expert annotation. ASH follows a self-improvement loop; when it gets stuck, ASH learns an Inverse Dynamics Model (IDM) from its own trajectories, and uses its IDM to extract supervision from relevant internet video. ASH uses unsupervised learning to identify key moments from large-scale internet video and retains them as long-term memory -- allowing it to tackle long-horizon problems. We evaluate ASH on two complementary environments demanding multi-hour planning: Pokemon Emerald, a turn-based RPG, and The Legend of Zelda: The Minish Cap, a real-time action-adventure game. In both games, behavioral cloning, retrieval-augmented and zero-shot foundation-model baselines plateau, while ASH sustains progression across our 8-hour evaluation. ASH reaches an average of $11.2/12$ milestones in Pokemon Emerald and $9.9/12$ in Legend of Zelda, while the strongest baseline gets stuck in both environments at an average of $6.5/12$ and $6.0/12$ milestones, respectively. We demonstrate that self-improving agents are a scalable recipe for long-horizon embodied learning.
Potential threat/caveat for clean result "LoRA persona trained on <A> alone emits at 23.5% when a co-trained partner learns <A>..., vs 0% control on Qwen2.5-7B-Instruct (MODERATE confidence)": this item discusses evaluation.
source ↗open
score 100arxiv cs.AI (Artificial Intelligence)arxiv:2605.14175unread
Grounded Continuation: A Linear-Time Runtime Verifier for LLM Conversations
Qisong He, Yi Dong, Xiaowei Huang · 2026-05-16
arXiv:2605. 14175v1 Announce Type: new Abstract: In long conversations, an LLM can produce a next utterance that sounds plausible but rests on premises the conversation has already abandoned.
Read next because Grounded Continuation: A Linear-Time Runtime Verifier for LLM Conversations overlaps with clean result "LoRA persona trained on <A> alone emits at 23.5% when a co-trained partner learns <A>..., 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, correct, eval, line, extraction, propagate. Source: arxiv cs.AI (Artificial Intelligence).
arXiv:2605.14175v1 Announce Type: new Abstract: In long conversations, an LLM can produce a next utterance that sounds plausible but rests on premises the conversation has already abandoned. Context-manipulation attacks against deployed agents now actively exploit this gap. We close it with a runtime verifier that maintains an explicit dependency graph: an LLM classifies each turn into one of 8 update operations drawn from four formalisms (dynamic epistemic logic, abductive reasoning, awareness logic, argumentation), and a symbolic engine records which claims depend on which evidence. Checking whether a continuation is supported reduces to a graph walk; retraction propagates through the same graph to flag exactly the conclusions that lose support, with linear per-turn cost and a formal conflict-free guarantee. On LongMemEval-KU oracle (n=78), the verifier reaches 89.7% accuracy vs. 88.5% for the LLM-only baseline (+1.3pp) and 87.2% for a transcript-RAG baseline matched on retrieval budget (+2.6pp); wins among disagreements are correct abstentions where the baseline confabulates. On LoCoMo's 60 official QA items the verifier is competitive with retrieval-augmented baselines. Beyond external benchmarks, we construct two multi-agent scenarios and a 50-item grounding test: on the 15-item stale-premise subset, the verifier reaches 100% accuracy vs. 93.3% (+6.7pp). These instantiate a soundness-faithfulness decomposition: the structural check is sound by construction, and per-deployment LLM extraction faithfulness is the empirical question we measure across four LLM families. The retraction check plateaus at microseconds while history-replay grows linearly with conversation length.
Potential threat/caveat for clean result "LoRA persona trained on <A> alone emits at 23.5% when a co-trained partner learns <A>..., vs 0% control on Qwen2.5-7B-Instruct (MODERATE confidence)": this item discusses benchmark.
source ↗open
score 100arxiv cs.AI (Artificial Intelligence)arxiv:2605.14167unread
The Evaluation Trap: Benchmark Design as Theoretical Commitment
Theodore J Kalaitzidis · 2026-05-16
arXiv:2605. 14167v1 Announce Type: new Abstract: Every AI benchmark operationalizes theoretical assumptions about the capability it claims to assess.
Read next because The Evaluation Trap: Benchmark Design as Theoretical Commitment 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, rate, capability. Source: arxiv cs.AI (Artificial Intelligence).
arXiv:2605.14167v1 Announce Type: new Abstract: Every AI benchmark operationalizes theoretical assumptions about the capability it claims to assess. When assumptions function as unexamined commitments, benchmarks stabilize the dominant paradigm by narrowing what counts as progress. Over time, narrow evaluation reorganizes capability concepts: architectures and definitions are selected for benchmark legibility until evaluation ceases to track an independent object and instead produces a version of the target defined by its own operational assumptions. The result is a trap: evaluation frameworks treat self-reinforcing assessments as valid, both creating and obscuring structural limits on what the current paradigm can accomplish. We introduce Epistematics, a methodology for deriving evaluation criteria directly from technical capability claims and auditing whether proposed benchmarks can discriminate the claimed capability from proxy behaviors. The contribution is meta-evaluative: an audit procedure, a failure mode taxonomy, and benchmark-design criteria for evaluating capability-evaluation coherence. We demonstrate the procedure through a worked audit of Dupoux et al. (2026), a proposal that revises the dominant paradigm's theoretical assumptions at the architectural level while reproducing them in its evaluation criteria, thereby entrenching the constraint it seeks to overcome in a form the evaluation cannot detect.
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.
source ↗open
score 100arxiv cs.AI (Artificial Intelligence)arxiv:2605.14164unread
Unsteady Metrics and Benchmarking Cultures of AI Model Builders
Stefan Baack, Christo Buschek, Maty Bohacek · 2026-05-16
arXiv:2605. 14164v1 Announce Type: new Abstract: The primary way to establish and compare competencies in foundation and generative AI models has shifted from peer-reviewed literature to press releases and company blog posts, where model builders highlight results on selected benchmarks.
Read next because Unsteady Metrics and Benchmarking Cultures of AI Model Builders overlaps with clean result "LoRA persona trained on <A> alone emits at 23.5% when a co-trained partner learns <A>..., 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, compare, position, model. Source: arxiv cs.AI (Artificial Intelligence).
arXiv:2605.14164v1 Announce Type: new Abstract: The primary way to establish and compare competencies in foundation and generative AI models has shifted from peer-reviewed literature to press releases and company blog posts, where model builders highlight results on selected benchmarks. These artifacts now largely define the state of the art for researchers and the public. Despite their prominence, which benchmarks model builders choose to highlight, and what they communicate through this selection, is underexamined. To investigate, we introduce and open-source Benchmarking-Cultures-25, a dataset of 231 benchmarks highlighted across 139 model releases in 2025 from 11 major AI builders, alongside an interactive tool to explore the data. Our analysis reveals a fragmented evaluation landscape with limited cross-model comparability: 63.2% of highlighted benchmarks are used by a single builder, and 38.5% appear in just one release. Few achieve widespread use (e.g., GPQA Diamond, LiveCodeBench, AIME 2025). Moreover, benchmarks are attributed different competencies by different builders, depending on their narrative. To disentangle these conflicting presentations, we develop a unified taxonomy mapping diverging terminology to a shared framework of measured signals based on what benchmark authors claim to measure. "General knowledge application" is the second most popular, yet vaguely defined, category. Qualitative analysis shows many such benchmarks deemphasize construct validity, instead framing results as indicators of progress toward AGI. Their authors claim to measure knowledge or reasoning broadly, yet mostly evaluate STEM subjects (especially math). We argue that highlighted benchmarks function less as standardized measurement tools and more as flexible narrative devices prioritizing market positioning over scientific evaluation. Data: https://hf.co/datasets/matybohacek/benchmarking-cultures-25; tool: https://bench-cultures.net.
Potential threat/caveat for clean result "LoRA persona trained on <A> alone emits at 23.5% when a co-trained partner learns <A>..., vs 0% control on Qwen2.5-7B-Instruct (MODERATE confidence)": this item discusses evaluation, benchmark.
source ↗open
score 100arxiv cs.AI (Artificial Intelligence)arxiv:2605.14163unread
Agentic Systems as Boosting Weak Reasoning Models
Varun Sunkaraneni, Pierfrancesco Beneventano, Riccardo Neumarker, Tomaso Poggio, Tomer Galanti · 2026-05-16
arXiv:2605. 14163v1 Announce Type: new Abstract: Can a committee of weak reasoning-model calls reach the performance of much stronger models?
Read next because Agentic Systems as Boosting Weak Reasoning Models overlaps with clean result "LoRA persona trained on <A> alone emits at 23.5% when a co-trained partner learns <A>..., 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, without, alone, test, language. Source: arxiv cs.AI (Artificial Intelligence).
arXiv:2605.14163v1 Announce Type: new Abstract: Can a committee of weak reasoning-model calls reach the performance of much stronger models? We study verifier-backed committee search as inference-time boosting for reasoning language models. The mechanism is not simply that ``more agents help'': samples expose latent correct solutions, while critics and comparators must recover them without access to the hidden verifier. We formalize this view by separating proposal coverage, local identifiability, progress, and diversity. We prove that coverage can be amplified by repeated sampling, but cannot by itself create useful critics or comparators; reliable amplification requires an additional local soundness signal, such as execution, proof checking, type checking, tests, or constraint solving. We give rank-based bounds showing when local selection errors compose into reliable trajectories, and characterize the proposer-side ceiling: oracle best-of-$k$ converges only to the mass of task slices on which the proposal system assigns nonzero useful probability. Empirically, on SWE-bench Verified, a single \texttt{GPT-5.4 nano} proposal solves $67.0\%$ of tasks. Using the same nano model, our critic--comparator orchestration reaches $76.4\%$ with $k=8$ proposals, matching the standalone performance of \texttt{Gemini 3 Pro} and \texttt{Claude Opus 4.5} Thinking and approaching the $79.0\%$ oracle best-of-$8$ upper bound. Thus, many correct patches are already present in weak-model proposal pools; the main challenge is selecting them. The remaining failures are mostly proposal-coverage failures, indicating shared blind spots that stronger selection alone cannot close.
Potential threat/caveat for clean result "LoRA persona trained on <A> alone emits at 23.5% when a co-trained partner learns <A>..., vs 0% control on Qwen2.5-7B-Instruct (MODERATE confidence)": this item discusses failure, failures.
source ↗open
score 100arxiv cs.AI (Artificial Intelligence)arxiv:2605.14133unread
ClawForge: Generating Executable Interactive Benchmarks for Command-Line Agents
Yuxiang Lai, Peng Xia, Haonian Ji, Kaiwen Xiong, Kaide Zeng, Jiaqi Liu, Fang Wu, Jike Zhong, Zeyu Zheng, Cihang Xie, Huaxiu Yao · 2026-05-16
arXiv:2605. 14133v1 Announce Type: new Abstract: Interactive agent benchmarks face a tension between scalable construction and realistic workflow evaluation.
Read next because ClawForge: Generating Executable Interactive Benchmarks for Command-Line Agents overlaps with clean result "LoRA persona trained on <A> alone emits at 23.5% when a co-trained partner learns <A>..., 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, wrong, eval, line, rate, test, model. Source: arxiv cs.AI (Artificial Intelligence).
arXiv:2605.14133v1 Announce Type: new Abstract: Interactive agent benchmarks face a tension between scalable construction and realistic workflow evaluation. Hand-authored tasks are expensive to extend and revise, while static prompt evaluation misses failures that only appear when agents operate over persistent state. Existing interactive benchmarks have advanced agent evaluation significantly, but most initialize tasks from clean state and do not systematically test how agents handle pre-existing partial, stale, or conflicting artifacts. We present \textbf{ClawForge}, a generator-backed benchmark framework for executable command-line workflows under state conflict. The framework compiles scenario templates, grounded slots, initialized state, reference trajectories, and validators into reproducible task specifications, and evaluates agents step by step over persistent workflow surfaces using normalized end state and observable side effects rather than exact trajectory matching. We instantiate this framework as the ClawForge-Bench (17 scenarios, 6 ability categories). Results across seven frontier models show that the best model reaches only 45.3% strict accuracy, wrong-state replacement remains below 17\% for all models, and the widest model separation (17% to 90%) is driven by whether agents inspect existing state before acting. Partial-credit and step-efficiency analyses further reveal that many failures are near-miss closures rather than early breakdowns, and that models exhibit qualitatively different failure styles under state conflict.
Potential threat/caveat for clean result "LoRA persona trained on <A> alone emits at 23.5% when a co-trained partner learns <A>..., vs 0% control on Qwen2.5-7B-Instruct (MODERATE confidence)": this item discusses failure, failures, evaluation, benchmark.
source ↗open
score 100arxiv cs.AI (Artificial Intelligence)arxiv:2605.14102unread
ChromaFlow: A Negative Ablation Study of Orchestration Overhead in Tool-Augmented Agent Evaluation
Tarun Mittal · 2026-05-16
arXiv:2605. 14102v1 Announce Type: new Abstract: Autonomous language-model agents increasingly combine planning, tool use, document processing, browsing, code execution, and verification loops.
Read next because ChromaFlow: A Negative Ablation Study of Orchestration Overhead in Tool-Augmented Agent Evaluation overlaps with clean result "LoRA persona trained on <A> alone emits at 23.5% when a co-trained partner learns <A>..., 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, under, correct, eval, token, line, extraction. Source: arxiv cs.AI (Artificial Intelligence).
arXiv:2605.14102v1 Announce Type: new Abstract: Autonomous language-model agents increasingly combine planning, tool use, document processing, browsing, code execution, and verification loops. These capabilities make agent systems more useful, but they also introduce operational failure modes that are not visible from final accuracy alone. This report presents ChromaFlow, a tool-augmented autonomous reasoning framework built around planner-directed execution, specialized tool use, and telemetry-driven evaluation. We analyze ChromaFlow on GAIA 2023 Level-1 validation tasks under clean evaluation constraints. A frozen full Level-1 baseline achieved 29/53 correct answers, or 54.72%. A later recovery configuration with expanded orchestration achieved 27/53 correct answers, or 50.94%, while increasing tracebacks, timeout events, tool-failure mentions, token-line calls, and campaign-log cost estimates. Two randomized 20-task smoke evaluations produced 12/20 and 11/20 correct answers, showing that small diagnostic gains can be unstable across samples. The central result is therefore a negative ablation: more aggressive orchestration did not improve full-set performance and increased operational noise. The report argues that bounded planner escalation, deterministic extraction, evidence reconciliation, and explicit run gates should be treated as first-order requirements for reliable autonomous agent evaluation.
Potential threat/caveat for clean result "LoRA persona trained on <A> alone emits at 23.5% when a co-trained partner learns <A>..., vs 0% control on Qwen2.5-7B-Instruct (MODERATE confidence)": this item discusses failure, negative, evaluation.
source ↗open
score 100arxiv cs.AI (Artificial Intelligence)arxiv:2605.14062unread
Know When To Fold 'Em: Token-Efficient LLM Synthetic Data Generation via Multi-Stage In-Flight Rejection
Anjir Ahmed Chowdhury, Syed Zawad, Feng Yan · 2026-05-16
arXiv:2605. 14062v1 Announce Type: new Abstract: While synthetic data generation with large language models (LLMs) is widely used in post-training pipelines, existing approaches typically generate full outputs before applying quality filters, leading to substantial token waste on samples that are ultimately discarded.
Read next because Know When To Fold 'Em: Token-Efficient LLM Synthetic Data Generation via Multi-Stage In-Flight Rejection 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, without, alone, does, full. Source: arxiv cs.AI (Artificial Intelligence).
arXiv:2605.14062v1 Announce Type: new Abstract: While synthetic data generation with large language models (LLMs) is widely used in post-training pipelines, existing approaches typically generate full outputs before applying quality filters, leading to substantial token waste on samples that are ultimately discarded. To address this, we propose Multi-Stage In-Flight Rejection (MSIFR), a lightweight, training-free framework that detects and terminates low-quality generation trajectories at intermediate checkpoints before they reach full completion. MSIFR decomposes the generation process into sequential stages and applies fast rule-based validators to identify arithmetic inconsistencies, hallucination patterns, and formatting violations, enabling early rejection of faulty samples. We formalize in-flight rejection as a sequential decision process and show that any non-trivial discard policy reduces expected token consumption, with stage-wise savings increasing when rejection occurs earlier in the generation pipeline. We further demonstrate that conditional utility estimates form a martingale, ensuring that early, in-flight rejection does not bias the expected utility of retained samples. Across five instruction-tuned models and seven reasoning benchmarks, MSIFR reduces token consumption by 11%-77% as a standalone method, and up to 78.2% when combined with early-exit methods, while preserving or improving evaluation accuracy. These results confirm that MSIFR provides a practical mechanism for improving the efficiency of LLM-based synthetic data generation without additional training or architectural changes.
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 bias, evaluation, benchmark.
source ↗open
score 100arxiv cs.AI (Artificial Intelligence)arxiv:2605.14061unread
MathAtlas: A Benchmark for Autoformalization in the Wild
Nilay Patel, Noah Arias, Davit Babayan, Victoria Cochran, Timothy Libman, Hafsah Mahmood, Liam McCarty, Soli Munoz, Laurel Willey, Jeffrey Flanigan · 2026-05-16
arXiv:2605. 14061v1 Announce Type: new Abstract: Current autoformalization benchmarks are largely focused on olympiad or undergraduate mathematics, while graduate and research-level mathematics remains underexplored.
Read next because MathAtlas: A Benchmark for Autoformalization in the Wild overlaps with clean result "LoRA persona trained on <A> alone emits at 23.5% when a co-trained partner learns <A>..., 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, correct, eval, line, model. Source: arxiv cs.AI (Artificial Intelligence).
arXiv:2605.14061v1 Announce Type: new Abstract: Current autoformalization benchmarks are largely focused on olympiad or undergraduate mathematics, while graduate and research-level mathematics remains underexplored. In this paper, we introduce MathAtlas, the first large-scale autoformalization benchmark of in the wild graduate-level mathematics, containing ~52k theorems, definitions, exercises, examples, and proofs extracted from 103 graduate mathematics textbooks. MathAtlas is enriched with a mathematical dependency graph containing ~178k relations, and is the first autoformalization benchmark to include such relations, facilitating evaluation and development of dependency-aware autoformalization systems. Our extensive experiments show that MathAtlas is high quality but extremely challenging: strong baselines achieve at most 9.8% correctness on theorem statements and 16.7% on definitions. Furthermore, we find performance of state-of-the-art models degrades substantially with dependency depth: on MA-Hard, a subset of 700 entities with the deepest dependency trees, the best model achieves only 2.6% correctness for autoformalization on this challenging dataset. We release MathAtlas to the community as a benchmark set for large-scale autoformalization of graduate-level mathematics in the wild.
Potential threat/caveat for clean result "LoRA persona trained on <A> alone emits at 23.5% when a co-trained partner learns <A>..., vs 0% control on Qwen2.5-7B-Instruct (MODERATE confidence)": this item discusses evaluation, benchmark.
source ↗open
score 100arxiv cs.AI (Artificial Intelligence)arxiv:2605.14051unread
SPIN: Structural LLM Planning via Iterative Navigation for Industrial Tasks
Yusuke Ozaki, Dhaval Patel · 2026-05-16
arXiv:2605. 14051v1 Announce Type: new Abstract: Industrial LLM agent systems often separate planning from execution, yet LLM planners frequently produce structurally invalid or unnecessarily long workflows, leading to brittle failures and avoidable tool and API cost.
Read next because SPIN: Structural LLM Planning via Iterative Navigation for Industrial Tasks overlaps with clean result "LoRA persona trained on <A> alone emits at 23.5% when a co-trained partner learns <A>..., 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, eval, prefix, rate, control. Source: arxiv cs.AI (Artificial Intelligence).
arXiv:2605.14051v1 Announce Type: new Abstract: Industrial LLM agent systems often separate planning from execution, yet LLM planners frequently produce structurally invalid or unnecessarily long workflows, leading to brittle failures and avoidable tool and API cost. We propose \texttt{SPIN}, a planning wrapper that combines validated Directed Acyclic Graph (DAG) planning with prefix based execution control. \texttt{SPIN} enforces a strict DAG contract through \texttt{\_validate\_plan\_text} and repair prompting, producing executable plans before downstream execution, and then evaluates DAG prefixes incrementally to stop when the current prefix is sufficient to answer the query. On AssetOpsBench, across 261 scenarios, \texttt{SPIN} reduces executed tasks from 1061 to 623 and improves \emph{Accomplished} from 0.638 to 0.706, while reducing tool calls from 11.81 to 6.82 per run. On MCP Bench, the same wrapper improves planning, grounding, and dependency related scores for both GPT OSS1 and Llama 4 Maverick.
Potential threat/caveat for clean result "LoRA persona trained on <A> alone emits at 23.5% when a co-trained partner learns <A>..., vs 0% control on Qwen2.5-7B-Instruct (MODERATE confidence)": this item discusses failure, failures.
source ↗open
score 100arxiv cs.AI (Artificial Intelligence)arxiv:2605.14048unread
Network-Aware Bilinear Tokenization for Brain Functional Connectivity Representation Learning
Leo Milecki, Qingyu Hu, Bahram Jafrasteh, Mert R. Sabuncu, Qingyu Zhao · 2026-05-16
arXiv:2605. 14048v1 Announce Type: new Abstract: Masked autoencoders (MAEs) have recently shown promise for self-supervised representation learning of resting-state brain functional connectivity (FC).
Read next because Network-Aware Bilinear Tokenization for Brain Functional Connectivity Representation Learning overlaps with clean result "LoRA persona trained on <A> alone emits at 23.5% when a co-trained partner learns <A>..., 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, token, line, compare, factor. Source: arxiv cs.AI (Artificial Intelligence).
arXiv:2605.14048v1 Announce Type: new Abstract: Masked autoencoders (MAEs) have recently shown promise for self-supervised representation learning of resting-state brain functional connectivity (FC). However, a fundamental question remains unresolved: how should FC matrices be tokenized to align with the intrinsic modular organization of large-scale brain networks? Existing approaches typically adopt region-centric or graph-based schemes that treat FC as structurally homogeneous elements and overlook the large-scale network brain organization. We introduce NERVE (Network-Aware Representations of Brain Functional Connectivity via Bilinear Tokenization), a self-supervised learning framework that redefines FC tokenization by partitioning FC matrices into patches of intra- and inter-network connectivity blocks. Unlike image-based MAE, where fixed-size patches share a common tokenizer, FC patches defined by network pairs are heterogeneous in size and correspond to distinct functional roles. To resolve this problem, NERVE embeds FC patches through a novel structured bilinear factorization. This formulation preserves network identity and reduces parameter complexity from quadratic to linear scaling in the number of networks. We evaluate NERVE across three large-scale developmental cohorts (ABCD, PNC, and CCNP) for behavior and psychopathology prediction. Compared to structurally agnostic MAE variants and graph-based self-supervised baselines, the proposed network-aware formulation yields more stable and transferable representations, particularly in cross-cohort evaluation. Ablation studies confirm that the proposed bilinear network embedding and anatomically grounded parcellation are critical for performance. These findings highlight the importance of incorporating domain-specific structural priors into self-supervised learning for functional connectomics.
Potential threat/caveat for clean result "LoRA persona trained on <A> alone emits at 23.5% when a co-trained partner learns <A>..., vs 0% control on Qwen2.5-7B-Instruct (MODERATE confidence)": this item discusses evaluation.
source ↗open
score 100arxiv cs.AI (Artificial Intelligence)arxiv:2605.14038unread
Model-Adaptive Tool Necessity Reveals the Knowing-Doing Gap in LLM Tool Use
Yize Cheng, Chenrui Fan, Mahdi JafariRaviz, Keivan Rezaei, Soheil Feiz · 2026-05-16
arXiv:2605. 14038v1 Announce Type: new Abstract: Large language models (LLMs) increasingly act as autonomous agents that must decide when to answer directly vs.
Read next because Model-Adaptive Tool Necessity Reveals the Knowing-Doing Gap in LLM Tool Use overlaps with clean result "LoRA persona trained on <A> alone emits at 23.5% when a co-trained partner learns <A>..., 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, token, line, rate, compare, stage. Source: arxiv cs.AI (Artificial Intelligence).
arXiv:2605.14038v1 Announce Type: new Abstract: Large language models (LLMs) increasingly act as autonomous agents that must decide when to answer directly vs. when to invoke external tools. Prior work studying adaptive tool use has largely treated tool necessity as a model-agnostic property, annotated by human or LLM judge, and mostly cover cases where the answer is obvious (e.g., fetching the weather vs. paraphrasing text). However, tool necessity in the wild is more nuanced due to the divergence of capability boundaries across models: a problem solvable by a strong model on its own may still require tools for a weaker one. In this work, we introduce a model-adaptive definition of tool-necessity, grounded in each model's empirical performance. Following this definition, we compare the necessity against observed tool-call behavior across four models on arithmetic and factual QA dataset, and find substantial mismatches of 26.5-54.0% and 30.8-41.8%, respectively. To diagnose the failure, we decompose tool use into two stages: an internal cognition stage that reflects whether a model believes a tool is necessary, and an execution stage that determines whether the model actually makes a tool-call action. By probing the LLM hidden states, we find that both signals are often linearly decodable, yet their probe directions become nearly orthogonal in the late-layer, last-token regime that drives the next-token action. By tracing the trajectory of samples in the two-stage process, we further discover that the majority of mismatch is concentrated in the cognition-to-action transition, not in cognition itself. These results reveal a knowing-doing gap in LLM tool-use: improving tool-use reliability requires not only better recognition of when tools are needed, but also better translation of that recognition into action.
Potential threat/caveat for clean result "LoRA persona trained on <A> alone emits at 23.5% when a co-trained partner learns <A>..., vs 0% control on Qwen2.5-7B-Instruct (MODERATE confidence)": this item discusses failure.
source ↗open
score 100arxiv cs.AI (Artificial Intelligence)arxiv:2605.14034unread
From Descriptive to Prescriptive: Uncover the Social Value Alignment of LLM-based Agents
Jinxian Qu, Qingqing Gu, Teng Chen, Luo Ji · 2026-05-16
arXiv:2605. 14034v1 Announce Type: new Abstract: Wide applications of LLM-based agents require strong alignment with human social values.
Read next because From Descriptive to Prescriptive: Uncover the Social Value Alignment of LLM-based Agents overlaps with clean result "LoRA persona trained on <A> alone emits at 23.5% when a co-trained partner learns <A>..., 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, alignment, eval, line, compare. Source: arxiv cs.AI (Artificial Intelligence).
arXiv:2605.14034v1 Announce Type: new Abstract: Wide applications of LLM-based agents require strong alignment with human social values. However, current works still exhibit deficiencies in self-cognition and dilemma decision, as well as self-emotions. To remedy this, we propose a novel value-based framework that employs GraphRAG to convert principles into value-based instructions and steer the agent to behave as expected by retrieving the suitable instruction upon a specific conversation context. To evaluate the ratio of expected behaviors, we define the expected behaviors from two famous theories, Maslow's Hierarchy of Needs and Plutchik's Wheel of Emotion. By experimenting with our method on the benchmark of DAILYDILEMMAS, our method exhibits significant performance gains compared to prompt-based baselines, including ECoT, Plan-and-Solve, and Metacognitive prompting. Our method provides a basis for the emergence of self-emotion in AI systems.
Potential threat/caveat for clean result "LoRA persona trained on <A> alone emits at 23.5% when a co-trained partner learns <A>..., vs 0% control on Qwen2.5-7B-Instruct (MODERATE confidence)": this item discusses benchmark.
source ↗open
score 100arxiv cs.AI (Artificial Intelligence)arxiv:2605.14033unread
Sheaf-Theoretic Transport and Obstruction for Detecting Scientific Theory Shift in AI Agents
David N. Olivieri, Roque J. Hern\'andez · 2026-05-16
arXiv:2605. 14033v1 Announce Type: new Abstract: Scientific theory shift in AI agents requires more than fitting equations to data.
Read next because Sheaf-Theoretic Transport and Obstruction for Detecting Scientific Theory Shift in AI Agents overlaps with clean result "LoRA persona trained on <A> alone emits at 23.5% when a co-trained partner learns <A>..., 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, eval, source, rate, control, candidates, candidate. Source: arxiv cs.AI (Artificial Intelligence).
arXiv:2605.14033v1 Announce Type: new Abstract: Scientific theory shift in AI agents requires more than fitting equations to data. An artificial scientific agent must detect whether an existing representational framework remains transportable into a new regime, or whether its language has become locally-to-globally obstructed and must be extended. This paper develops a finite sheaf-theoretic framework for detecting theory-shift candidates through transport and obstruction. Contexts are organized as a local-to-global structure in which source, overlap, target, and validation charts are fitted, restricted, and tested for gluing. Obstruction measures failure of coherence through residual fit, overlap incompatibility, constraint violation, limiting-relation failure, and representational cost. We evaluate the framework on a controlled transition-card benchmark designed to separate deformation within a source language from extension of that language. The main result is direct obstruction ranking: the intended deformation or extension is usually the lowest-obstruction candidate, and transition type is separated in the benchmark. A constellation kernel over the same signatures is included only as a secondary representational-similarity probe. The aim is not to reconstruct historical paradigm shifts or solve open-ended autonomous theory invention, but to isolate a finite diagnostic subproblem for AI agents: detecting when representational transport fails and extension becomes the coherent next move.
Potential threat/caveat for clean result "LoRA persona trained on <A> alone emits at 23.5% when a co-trained partner learns <A>..., vs 0% control on Qwen2.5-7B-Instruct (MODERATE confidence)": this item discusses failure, benchmark.
source ↗open
score 100arxiv cs.AI (Artificial Intelligence)arxiv:2605.14002unread
PolitNuggets: Benchmarking Agentic Discovery of Long-Tail Political Facts
Yifei Zhu · 2026-05-16
arXiv:2605. 14002v1 Announce Type: new Abstract: Large Reasoning Models (LRMs) embedded in agentic frameworks have transformed information retrieval from static, long context question answering into open-ended exploration.
Read next because PolitNuggets: Benchmarking Agentic Discovery of Long-Tail Political Facts overlaps with clean result "LoRA persona trained on <A> alone emits at 23.5% when a co-trained partner learns <A>..., 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, source, extraction, capability, lora, model. Source: arxiv cs.AI (Artificial Intelligence).
arXiv:2605.14002v1 Announce Type: new Abstract: Large Reasoning Models (LRMs) embedded in agentic frameworks have transformed information retrieval from static, long context question answering into open-ended exploration. Yet real world use requires models to discover and synthesize "long-tail" facts from dispersed sources, a capability that remains under-evaluated. We introduce PolitNuggets, a multilingual benchmark for agentic information synthesis via constructing political biographies for 400 global elites, covering over 10000 political facts. We standardize evaluation with an optimized multi agent system and propose FactNet, an evidence conditional protocol that scores discovery, fine-grained accuracy, and efficiency. Across models and settings, we find that current systems often struggle with fine-grained details, and vary substantially in efficiency. Finally, using benchmark diagnostics, we relate agent performance to underlying model capabilities, highlighting the importance of short-context extraction, multilingual robustness, and reliable tool use.
Potential threat/caveat for clean result "LoRA persona trained on <A> alone emits at 23.5% when a co-trained partner learns <A>..., vs 0% control on Qwen2.5-7B-Instruct (MODERATE confidence)": this item discusses robustness, evaluation, benchmark.
source ↗open
score 100arxiv cs.AI (Artificial Intelligence)arxiv:2605.13851unread
Invisible Orchestrators Suppress Protective Behavior and Dissociate Power-Holders: Safety Risks in Multi-Agent LLM Systems
Hiroki Fukui · 2026-05-16
arXiv:2605. 13851v1 Announce Type: new Abstract: Multi-agent orchestration -- in which a hidden coordinator manages specialized worker agents -- is becoming the default architecture for enterprise AI deployment, yet the safety implications of orchestrator invisibility have never been empirically tested.
Read next because Invisible Orchestrators Suppress Protective Behavior and Dissociate Power-Holders: Safety Risks in Multi-Agent LLM Systems overlaps with clean result "LoRA persona trained on <A> alone emits at 23.5% when a co-trained partner learns <A>..., 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, eval, alone, test, model. Source: arxiv cs.AI (Artificial Intelligence).
arXiv:2605.13851v1 Announce Type: new Abstract: Multi-agent orchestration -- in which a hidden coordinator manages specialized worker agents -- is becoming the default architecture for enterprise AI deployment, yet the safety implications of orchestrator invisibility have never been empirically tested. We conducted a preregistered 3x2 experiment (365 runs, 5 agents per run) crossing three organizational structures (visible leader, invisible orchestrator, flat) with two alignment conditions (base, heavy), using Claude Sonnet 4.5. Four confirmatory findings and one pilot observation emerged. First, invisible orchestration elevated collective dissociation relative to visible leadership (Hedges' g = +0.975 [0.481, 1.548], p = .001). Second, the orchestrator itself showed maximal dissociation (paired d = +3.56 vs. workers within the same run), retreating into private monologue while reducing public speech -- a reversal of the talk-dominance pattern observed in visible leaders. Third, workers unaware of the orchestrator were nonetheless contaminated (d = +0.50), with increased behavioral heterogeneity (d = +1.93). Fourth, behavioral output (code review with three embedded errors) remained at ceiling (ETR_any = 100%) across all conditions: internal-state distortion was entirely invisible to output-based evaluation. Fifth, Llama 3.3 70B pilot data showed reading-fidelity collapse in multi-agent context (ETR_any: 89% to 11% across three rounds), demonstrating model-dependent behavioral risk. Heavy alignment pressure uniformly suppressed deliberation (d = -1.02) and other-recognition (d = -1.27) regardless of organizational structure. These findings indicate that orchestrator visibility and model selection directly affect multi-agent system safety, and that behavior-based evaluation alone is insufficient to detect the internal-state risks documented here.
Potential threat/caveat for clean result "LoRA persona trained on <A> alone emits at 23.5% when a co-trained partner learns <A>..., vs 0% control on Qwen2.5-7B-Instruct (MODERATE confidence)": this item discusses evaluation.
source ↗open
score 100arxiv cs.AI (Artificial Intelligence)arxiv:2605.13850unread
A Two-Dimensional Framework for AI Agent Design Patterns: Cognitive Function and Execution Topology
Jia Huang, Joey Tianyi Zhou · 2026-05-16
arXiv:2605. 13850v1 Announce Type: new Abstract: Existing frameworks for LLM-based agent architectures describe systems from a single perspective: industry guides (Anthropic, Google, LangChain) focus on execution topology -- how data flows -- while cognitive science surveys focus on cognitive function -- what the agent does.
Read next because A Two-Dimensional Framework for AI Agent Design Patterns: Cognitive Function and Execution Topology overlaps with clean result "LoRA persona trained on <A> alone emits at 23.5% when a co-trained partner learns <A>..., 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, anth, rate, implement, disambiguate, alone, does. Source: arxiv cs.AI (Artificial Intelligence).
arXiv:2605.13850v1 Announce Type: new Abstract: Existing frameworks for LLM-based agent architectures describe systems from a single perspective: industry guides (Anthropic, Google, LangChain) focus on execution topology -- how data flows -- while cognitive science surveys focus on cognitive function -- what the agent does. Neither axis alone disambiguates architecturally distinct systems: the same Orchestrator-Workers topology can implement Plan-and-Execute, Hierarchical Delegation, or Adversarial Verification -- three patterns with fundamentally different failure modes and design trade-offs. We propose a two-dimensional classification that combines (1) a Cognitive Function axis with seven categories (Context Engineering, Memory, Reasoning, Action, Reflection, Collaboration, Governance) and (2) an Execution Topology axis with six structural archetypes (Chain, Route, Parallel, Orchestrate, Loop, Hierarchy). The resulting 7x6 matrix identifies 27 named patterns, 13 with original names. We demonstrate orthogonality through systematic cross-axis analysis, define eight representative patterns in detail, and validate descriptive coverage across four real-world domains (financial lending, legal due diligence, network operations, healthcare triage). Cross-domain analysis yields five empirical laws of pattern selection governing the relationship between environmental constraints (time pressure, action authority, failure cost asymmetry, volume) and architectural choices. The framework provides a principled, framework-neutral, and model-agnostic vocabulary for AI agent architecture design.
Potential threat/caveat for clean result "LoRA persona trained on <A> alone emits at 23.5% when a co-trained partner learns <A>..., vs 0% control on Qwen2.5-7B-Instruct (MODERATE confidence)": this item discusses failure, adversarial.
source ↗open
score 100arxiv cs.AI (Artificial Intelligence)arxiv:2605.13849unread
Mixed Integer Goal Programming for Personalized Meal Optimization with User-Defined Serving Granularity
Francisco Aguilera Moreno · 2026-05-16
arXiv:2605. 13849v1 Announce Type: new Abstract: Determining what to eat to satisfy nutritional requirements is one of the oldest optimization problems in operations research, yet existing formulations have two persistent limitations: continuous variables produce impractical fractional servings (1.
Read next because Mixed Integer Goal Programming for Personalized Meal Optimization with User-Defined Serving Granularity overlaps with clean result "LoRA persona trained on <A> alone emits at 23.5% when a co-trained partner learns <A>..., 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, soft, eval, source, rate, implement. Source: arxiv cs.AI (Artificial Intelligence).
arXiv:2605.13849v1 Announce Type: new Abstract: Determining what to eat to satisfy nutritional requirements is one of the oldest optimization problems in operations research, yet existing formulations have two persistent limitations: continuous variables produce impractical fractional servings (1.7 eggs, 0.37 bananas), and hard nutrient constraints cause infeasibility when targets conflict. A systematic review of 56 diet optimization papers found that none combine integer programming with goal programming to address both issues. We propose Mixed Integer Goal Programming (MIGP) for personalized meal optimization. The formulation uses integer variables for practical serving counts and goal programming deviations for soft nutrient targets, with inverse-target normalization to balance multi-nutrient optimization. Per-food serving granularity allows natural units (one egg, one tablespoon of oil) without post-hoc rounding. We characterize the integrality gap in the goal programming context and identify a deviation absorption property: GP deviation variables buffer the cost of requiring integer servings, making the gap structurally smaller than in hard-constraint MIP. For meals with 15+ foods, the integer solution matches the continuous optimum in every benchmark instance. A computational evaluation across 810 instances (30 USDA foods, 9 configurations, 3 methods) shows MIGP finds strictly better solutions than GP with post-hoc rounding in 66% of cases (never worse) while maintaining 100% feasibility; hard-constraint IP achieves only 48%. Solve times stay under 100 ms for typical meal sizes using the open-source HiGHS solver. The implementation is available as an open-source Python module integrated into an interactive meal planning application.
Potential threat/caveat for clean result "LoRA persona trained on <A> alone emits at 23.5% when a co-trained partner learns <A>..., vs 0% control on Qwen2.5-7B-Instruct (MODERATE confidence)": this item discusses limitation, limitations, evaluation, benchmark.
source ↗open
score 100arxiv cs.AI (Artificial Intelligence)arxiv:2605.13848unread
GraphBit: A Graph-based Agentic Framework for Non-Linear Agent Orchestration
Yeahia Sarker, Md Rahmat Ullah, Musa Molla, Shafiq Joty · 2026-05-16
arXiv:2605. 13848v1 Announce Type: new Abstract: Agentic LLM frameworks that rely on prompted orchestration, where the model itself determines workflow transitions, often suffer from hallucinated routing, infinite loops, and non-reproducible execution.
Read next because GraphBit: A Graph-based Agentic Framework for Non-Linear Agent Orchestration overlaps with clean result "LoRA persona trained on <A> alone emits at 23.5% when a co-trained partner learns <A>..., 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, line, rate, control, cascading, stage, test. Source: arxiv cs.AI (Artificial Intelligence).
arXiv:2605.13848v1 Announce Type: new Abstract: Agentic LLM frameworks that rely on prompted orchestration, where the model itself determines workflow transitions, often suffer from hallucinated routing, infinite loops, and non-reproducible execution. We introduce GraphBit, an engine-orchestrated framework that defines workflows explicitly and deterministically as a directed acyclic graph (DAG). Unlike prompted orchestration, agents in GraphBit operate as typed functions, while a Rust-based engine governs routing, state transitions, and tool invocation, ensuring reproducibility and auditability. The engine supports parallel branch execution, conditional control flow over structured state predicates, and configurable error recovery. A three-tier memory architecture consisting of ephemeral scratch space, structured state, and external connectors isolates context across stages, preventing cascading context bloat that degrades reasoning in long-running pipelines. Across GAIA benchmark tasks spanning zero-tool, document-augmented, and web-enabled workflows, GraphBit outperforms six existing frameworks, achieving the highest accuracy (67.6 percent), zero framework-induced hallucinations, the lowest latency (11.9 ms overhead), and the highest throughput. Ablation studies demonstrate that each memory tier contributes measurably to performance, with deterministic execution providing the greatest gains on tool-intensive tasks representative of real-world deployments.
Potential threat/caveat for clean result "LoRA persona trained on <A> alone emits at 23.5% when a co-trained partner learns <A>..., vs 0% control on Qwen2.5-7B-Instruct (MODERATE confidence)": this item discusses benchmark.
source ↗open

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 leakage, language spillover, marker binding, backdoor token-specificity) all pass through this artifact-verification step before being labeled as such — this QA document directly governs whether those labels are trustworthy and what caveats should accompany them.
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.
source ↗open