Paper

Drifting Objectives for Refining Discrete Diffusion Language Models

Unreadunread

AI summary

arXiv:2605. 19470v1 Announce Type: new Abstract: Discrete diffusion language models (DDLMs) generate text by iteratively denoising categorical token sequences, while recent drifting methods for continuous generators suggest that part of this sampling-time correction can instead be absorbed into training through an anti-symmetric fixed-point objective.

Reading status

Priority

About this paper

Authors: Daisuke Oba, Hiroki Furuta, Naoaki Okazaki
Released: 2026-05-20 (2mo ago)
Topic: Related to current project
Relation to my work: Read next because Drifting Objectives for Refining Discrete Diffusion 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, rect, correct, soft, token, line, rate, control. Source: arxiv cs.CL (NLP).
arXiv: 2605.19470

Ask Claude about this paper

Posts a comment with @claude so the runner spins up a Claude Code session scoped to Drifting Objectives for Refining Discrete Diffusion Language Models. Replies appear in the conversation below.

Comments

No comments yet.

Related records

No links yet.