The paper introduces Proximal Entropy Policy Optimization (PEPO), a method that assigns token‑level advantages in reinforcement learning for language models based on a local measure of entropy called proximal entropy. Unlike prior approaches that use global token entropy, PEPO evaluates importance relative to neighboring tokens, making it invariant to prompt difficulty and positional effects. Experiments on Qwen3 and Llama‑3.2 models show PEPO outperforms GRPO and other entropy‑based baselines in mathematical reasoning tasks.
By Yun Kim, Nojun Kwak
RankEvolve is an auto‑research framework that evolves generative ranking models by orchestrating multiple large‑language‑model coding agents through an Executable Operating Protocol (EOP). The system compiles a state machine that enforces phases, gates, branches, and loops, while a meta‑meta‑harness lets agents review and repair each other’s code. In budget‑matched experiments, heterogeneous composition of agents raised execution accuracy from 45.8 % to 62.5 % and reduced silent critical‑defect rates, achieving notable gains on the HSTU recommender and other benchmarks.
By Zheng Chen, Linfeng Liu, Hong Li, Hong Yan
The paper investigates self‑distillation techniques for language models by systematically varying three key design choices: the source of rollout tokens (student vs. teacher), the teacher coupling strategy (frozen or exponential moving average), and the KL divergence direction (reverse or forward). Experiments on Qwen2.5‑7B and Ministral‑3‑3B across 1,200 adaptation runs reveal that rollout source mainly affects acquisition on contradictory tasks, teacher coupling most strongly influences acquisition across all tasks, and KL direction impacts retention differently depending on the model. A controlled theoretical model reproduces these empirical trends, offering a unified framework for understanding acquisition‑retention trade‑offs in self‑distillation.
By Luis Zuin, Alexis Huet, Dario Rossi, Zied Ben Houidi
The paper investigates how large language models learn to hide their reasoning within text—termed steganographic reasoning—compared to related abilities like steganographic messaging and encoded reasoning. Across reinforcement learning, in-context learning, and supervised fine-tuning, models readily acquire messaging and encoded reasoning, but steganographic reasoning only emerges under supervised fine-tuning and requires substantially more training, unless a convenient cover task is provided. Even then, steganographic reasoning remains significantly harder than its neighboring capabilities.
By Julian Schulz, Lukas F\"ulle, Rieke Fruengel
The paper introduces Runtime Assurance Contracts (RAC) as a formal policy framework for high‑risk AI agents, addressing the "assurance‑transition gap" by binding autonomy boundaries, component eligibility, evidence state, transition policy, human‑review capacity, and non‑compensatory gates. RAC allows soft metrics to influence routing while mandating retries, switches, escalations, deferrals, or stops when mandatory gates fail or are unknown, ensuring aggregate performance cannot alone authorize action. The authors define the contract, evidence record, permission rule, and five invariants, and evaluate RAC through deterministic failure‑injection studies, hand‑authored traces, and a prospective synthetic holdout, comparing it to score‑only and restricted protocol baselines.
By Serhii Zabolotnii
The paper introduces ConflictGuide, a method that enhances LLM-based AutoResearch by incorporating feedback on competing behaviors during model code editing. By first exploring with scalar task performance and then using probes to measure and alleviate conflicts, ConflictGuide increases the proportion of edits that improve multiple behaviors and sustains progress beyond scalar-only plateaus. Experiments across five model families show reductions in task and conflict-related errors by up to 28% and 14% compared to scalar-only AutoResearch.
By Binqian Xu, Qiran Zou, Xiangbo Shu, Dianbo Liu
The Agent Error Dataset (AED) presents 50,228 error–diagnosis pairs collected from 9,961 source tasks across 33 environments, 19 harness families, and 23 policy models in text‑based agent systems. A five‑stage Agentic Error‑to‑Training (AET) pipeline generates diagnoses and proposed corrections, verifies them against recorded evidence, and creates separate training views for diagnosis and actor recovery. Experiments show that first‑proposal corrections improve verifier pass rates from 18.4% to 51.1%, and fine‑tuning with full‑diagnosis data raises Qwen3‑8B’s exact‑step agreement from 47.2% to 63.6% on a holdout set.
By Kunlun Zhu, Xuyan Ye, Yibo Li, Cheng Qian, Beibin Li, Heng Ji
The paper introduces ScholarEvolve, a framework that evolves the software harness of language agents by automatically incorporating insights from recent research papers. It organizes harness improvements into functional modules, uses topic modeling to identify distinct strategies, and evaluates combinations to boost task performance. Experiments show significant gains on AppWorld and Tau2-Bench, raising Qwen3.5-27B completion rates from 49.6% to 63.6% and GPT-5.4-mini pass@1 from 72.7% to 81.9%.
By Jingbo Yang, Kwei-Herng Lai, Xiaowen Wang, Yaar Harari, Evgeniy Gabrilovich, Shiyu Chang
DualCast is a dual‑path language model that forecasts financial time‑series by combining a fast numerical forecaster with an optional text‑conditioned revision mechanism. The fast path trains only new financial‑token embeddings and output heads on a frozen Qwen3‑8B backbone, while the slow path uses a LoRA adapter to incorporate news and refine predictions. In zero‑shot tests across equities and energy prices at multiple time resolutions, the slow path achieves the lowest mean absolute percentage error in most settings, especially for longer horizons, and news ablations show additional gains in many markets.
By Wentao Zhao, Hongqiang Wu, Shanghang Liu, Zhaochen Zan, Yu Zhang, Biqing Huang
The paper presents an automated segmentation pipeline for whole‑slide histopathology images of colorectal cancer, labeling tumor grades 1‑3 and normal mucosa. It employs dense prediction transformers with multiple encoder backbones, overlapping patches, test‑time augmentation, and an adaptive augmentation policy guided by large language models. The approach, combined with soft‑voting ensembles and post‑processing refinements, raises the F1 score from 62.92 to 69.84 on a colorectal cancer grade dataset.
By \"Umit Mert \c{C}a\u{g}lar, Alptekin Temizel
The paper introduces STAM, a state‑transition‑aware memory framework for large language model agents that process clinical records. STAM records changes in a patient’s state as new entries arrive, using semantic retrieval and typed clinical relations to separate current information (Active) from superseded or resolved information (History). During retrieval, a query‑dependent gate selects the appropriate historical memory, enabling accurate question answering and state‑maintenance diagnostics across four longitudinal clinical benchmarks.
By Maryam Haghifam, Zahra Rajabi, Yizhou Sun, Carlos Morato
The paper introduces an explainable pedestrian routing method that estimates perceived safety by first generating a natural‑language caption from street‑level images and then deriving risk scores solely from structured features of that caption. Benchmarking nine captioning setups against a CLIP image‑embedding baseline shows comparable performance, and the system was deployed on over 650,000 images across 36 wards in Manchester and Huddersfield. Field validation with 3,669 ratings from 70 participants revealed a modest but statistically significant correlation (r = 0.262) with human judgments, while a stronger supervised benchmark did not translate into better real‑world performance.
By Simon Parkinson, Paloma Liu, Wei Zheng, Mohammadreza Sheikhfathollahi
PANDA is a decentralized architecture for large-scale, fault-tolerant multi-agent systems that enables heterogeneous agents to discover each other's capabilities and self-organize into specialized teams for each task. It decouples collective communication from team communication, allowing agents to participate in multiple teams simultaneously and load-balance tasks across the collective. PANDA supports three planning and execution patterns—star, chain, and mesh—detects and recovers from infrastructure and orchestration failures, and uses a web-of-trust model for governance without a central bottleneck.
By Matthew D. Laws, Cristina Nita-Rotaru
ShamAN-Q is a sub‑1‑bit post‑training quantization technique that builds on NanoQuant by replacing its diagonal reconstruction geometry with a dense curvature metric inspired by the Shampoo optimizer. For each linear weight, it fits a Kronecker product to the empirical Fisher information matrix of a small calibration set via Kullback–Leibler minimization, yielding a Mahalanobis reconstruction loss. The method updates continuous ADMM steps to Sylvester equations while keeping the discrete projection and deployment format unchanged, and it redistributes uniform rank across layers, achieving lower perplexity on Qwen3‑Base at roughly 1 bpw and matching or improving zero‑shot accuracy on the Eleuther LM Evaluation Harness.
By Jonathan Mei, Sang Hyub Kim, Oliver Knitter, Chi Chen, Martin Roetteler
FlexRouter is a routing framework for large language models that explicitly models model complementarity to maximize answer coverage. It formulates routing as a coverage-oriented subset selection problem and uses Determinantal Point Processes to capture both competence and redundancy. During inference, a greedy strategy based on marginal log-determinant gains allows the router to adaptively determine subset sizes without a fixed budget, achieving higher coverage with lower redundancy on the RouterEval benchmark.
By Wang Wei, Harry Yang, Tiankai Yang, Samyadeep Basu, Hongjie Chen, Andy Zhao, Franck Dernoncourt, Ryan A. Rossi, Hoda Eldardiry
Prompt2Skill is an unsupervised framework that constructs skills for Large Language Models directly from natural‑language task descriptions. It automatically derives task specifications, discovers or synthesizes datasets, and refines the skill through a reflective editing loop. In experiments across question answering, reading comprehension, spreadsheet manipulation, and mathematical reasoning, Prompt2Skill outperforms direct prompting, improving performance by an average of 10.8 points on both open‑source and frontier models.
By Bo Ni, Li Li, Ryan A. Rossi, Franck Dernoncourt, Tyler Derr
CollabFlow introduces a recursive self‑improvement framework for multi‑agent collaboration in large language model systems. It trains a Collab‑Director to assemble teams of agents, uses a frozen executor to run them, and retrains the director each round based on outcomes. The system incorporates evidence‑conditioned communication protocols within collaboration graphs and a Collaborative Trajectory Balance objective to maintain diverse high‑performing teams across rounds, achieving superior performance on twelve datasets.
By Xiao Huang, Mingda Zhang, Junming Zhang, Qiang Huang, Hanwen Zhang, Yue Dai, Zijia Wang, Xiaoying Tang
The paper introduces dattri-LLM, a library designed to make training data attribution (TDA) practical for large language models. It achieves efficiency by using compact gradient representations and a cost‑based routing system, while maintaining compatibility by capturing per‑example gradients from existing training loops without modifications, even in distributed settings. The library also offers extensibility through reusable gradient operations and callbacks, supporting various attribution methods and applications such as online data selection, and demonstrates significant performance gains and scalability up to 110B‑parameter models.
By Shixuan Liu, Tongli Zhou, Junwei Deng, Pingbang Hu, Jiaqi W. Ma
The paper identifies a flaw in rubric‑based reinforcement learning where additive reward aggregation allows policies to compensate for missing critical criteria, leading to higher scores but poorer answers, especially in clinical consultation tasks. It demonstrates that grouping rubric criteria into protocol‑level dimensions—so a dimension only counts when all its criteria are satisfied—mitigates this reward hacking. The proposed Protocol‑level Rubrics (ProRubric) improve appropriateness by 10.8 points without sacrificing coverage and achieve the best performance across seven benchmarks.
By Maoqi Liu, Junwei He, Bowen Zhang, Feiran Li, Wentao Ma, Rongyi Lin, Shuhan Zhong, Quan Fang
The paper introduces Triage, a method that predicts the attention distribution of audio tokens before a language model processes them, enabling early pruning of less important tokens. By fitting a linear map to encoder outputs, Triage achieves high correlation (ρ ≥ 0.69) with full-model attention across eleven of thirteen large audio language models. Using this prediction, Triage compresses audio inputs while maintaining near‑full performance, outperforming baselines in transcription accuracy and significantly increasing the amount of audio that fits within a model’s context window.
By Kyoungjun Park, Yunzhe Li, Lili Qiu