The paper introduces Juris Policy Optimization (JPO), a post‑training framework designed to enhance structured legal reasoning in Chinese criminal judgment prediction. JPO first trains models with teacher‑generated rationales to guide a four‑step reasoning process, then applies reinforcement learning using a composite reward that balances prediction accuracy, reasoning completeness, and cross‑step consistency. Experiments on several open‑source language models and three Chinese legal benchmarks demonstrate that JPO consistently outperforms both supervised fine‑tuning and standard reinforcement learning baselines in terms of judgment prediction and reasoning quality.
By Zhaolu Kang, Yantao Liu, Tailong Luo, Leqi Zheng, Lei Wei, Chenghua Zhu, Junhao Gong, Jiachen Qian, Eric Hanchen Jiang, Jiaxin Liu, Yuan Wang, Hao Zhang, Zixia Wang, Rong Fu, Zheng Lin, Richeng Xuan, Zhichao Hu
The paper investigates on‑policy distillation (OPD), showing that teacher supervision during OPD contains significant noise that grows with teacher size, yet the student policy remains largely unaffected by this noise. It finds that OPD’s gains stem mainly from suppressing low‑log‑probability tokens, a process that can be replicated without a teacher. Building on this insight, the authors propose On‑Policy Self‑Adaptation (OPSA), a supervision‑free method that uses entropy‑adaptive negative advantages to improve performance on several benchmarks, outperforming both the base model and OPD.
By Yi Ding, Ruqi Zhang
The paper introduces X-Coder, a competitive programming model trained entirely on synthetic tasks, verified solutions, and reliable test cases, eliminating the need for real-world data in post‑training. A dual‑verification strategy is used to reduce noise in solutions and test outputs, providing high‑quality reward signals for reinforcement learning. X‑Coder‑14B achieves significant performance gains, scoring 67.5% on LiveCodeBench v5 and 63.4% on v6, surpassing its base model by over 40 points.
By Jie Wu, Haoling Li, Xin Zhang, Jiani Guo, Jane Luo, Xuewei Yang, Steven Liu, Yangyu Huang, Ruihang Chu, Scarlett Li, Yujiu Yang
arXiv:2608.30719v1 Announce Type: new
Abstract: Productive dialogue alignment requires distinguishing \emph{surface coordination} (acknowledgments and smooth task progression) from \emph{epistemic al...
By Yifan Zhu, Kyeongmin Rim, James Pustejovsky
AutoREC is an open‑source Python platform that uses reinforcement learning to automatically generate equivalent circuit models (ECMs) from electrochemical impedance spectroscopy (EIS) data. The platform frames ECM generation as a Markov decision process, training a Double Deep Q‑Network agent that iteratively modifies circuit topologies based on state, actions, and model feedback. It supports end‑to‑end workflows—including EIS preprocessing, agent training, ECM generation, and visualization—and has been demonstrated on synthetic datasets and real experimental spectra from batteries, corrosion, oxygen evolution, and CO₂ reduction systems.
By Ali Jaberi (Clean Energy Innovation Research Centre, National Research Council Canada, Mississauga, ON, Canada), Yonatan Kurniawan (Department of Materials Science and Engineering, University of Toronto, Toronto, ON, Canada), Robert Black (Clean Energy Innovation Research Centre, National Research Council Canada, Mississauga, ON, Canada), Shayan Mousavi M. (Clean Energy Innovation Research Centre, National Research Council Canada, Mississauga, ON, Canada), Kabir Verma (Cheriton School of Computer Science, University of Waterloo, Waterloo, ON, Canada), Zoya Sadighi (Clean Energy Innovation Research Centre, National Research Council Canada, Mississauga, ON, Canada), Santiago Miret (Lila Sciences, San Francisco, CA, USA), Jason Hattrick-Simpers (Department of Materials Science and Engineering, University of Toronto, Toronto, ON, Canada)
The paper introduces Adaptive Doubly Robust (ADR), an off‑policy evaluation method for ranking policies that blends adaptive importance weighting with reward regression to reduce variance. ADR is unbiased when the true user behavior model is known and, under a sufficient condition, achieves lower variance than the prior Adaptive Inverse Propensity Scoring (AIPS) approach. Experiments on synthetic data show that ADR consistently improves mean squared error over AIPS and other ranking OPE estimators across various data sizes and ranking lengths.
By Kosuke Iguchi, Ren Kishimoto
Titans-QFWP is a hybrid reinforcement learning architecture that combines a Quantum Fast Weight Programmer with a Titans-style memory system (Persistence, Surprise, and Forgetting) for adaptive portfolio optimization. It employs an enhanced A3C² framework with Hungarian-aligned K‑means clustering and scaled log‑return rewards to handle high‑dimensional market features. Tested on 468 S&P 500 stocks with about 3,000 trainable parameters, the model achieves strong performance metrics (median ARR 0.4260, Calmar 8.5504, IR 0.8427) and demonstrates that quantum gating reshapes memory roles to support drawdown control, return generation, and stabilization.
By Ming-Kai Hung, Jun-Hao Chen, Yun-Cheng Tsai, Samuel Yen-Chi Chen
The paper introduces Personalized Group Relative Policy Optimization (P‑GRPO), a new alignment framework for large language models that separates advantage estimation from immediate batch statistics. By normalizing advantages using preference‑group‑specific reward histories instead of the concurrent generation group, P‑GRPO maintains contrastive signals for distinct user preferences. Experiments across various tasks show that P‑GRPO converges faster and yields higher rewards than standard GRPO, improving alignment with heterogeneous human preferences while preserving general capabilities.
By Jialu Wang, Heinrich Peters, Asad A. Butt, Navid Hashemi, Alireza Hashemi, Pouya M. Ghari, Joseph Hoover, James Rae, Morteza Dehghani
arXiv:2608.29831v1 Announce Type: new
Abstract: Localizing issue-relevant code regions is a critical step in automated software engineering. However, due to their reliance on sparse trajectory-level...
By Doyeon Kim, Suyoung Bae, Yumin Lee, Jee-Hyong Lee
arXiv:2506.17007v3 Announce Type: replace
Abstract: A major bottleneck in scientific discovery consists of narrowing an exponentially large set of objects, such as proteins or molecules, to a small s...
By Marco Jiralerspong, Esther Derman, Danilo Vucetic, Esmeralda S. Whitammer, Bilun Sun, Tianyu Zhang, Pierre-Luc Bacon, Gauthier Gidel
arXiv:2310.02779v3 Announce Type: replace
Abstract: Generative flow networks (GFlowNets) are sequential sampling models trained to match a given distribution. GFlowNets have been successfully applied...
By Marco Jiralerspong, Bilun Sun, Danilo Vucetic, Tianyu Zhang, Yoshua Bengio, Gauthier Gidel, Esmeralda S. Whitammer
arXiv:2210.00580v4 Announce Type: replace
Abstract: This paper builds bridges between two families of probabilistic algorithms: (hierarchical) variational inference (VI), which is typically used to m...
By Esmeralda S. Whitammer, Salem Lahlou, Tristan Deleu, Xu Ji, Edward Hu, Katie Everett, Dinghuai Zhang, Yoshua Bengio
arXiv:2608.30910v1 Announce Type: new
Abstract: Spectroscopic structure elucidation is central to molecular analysis, but recent Large Language Model (LLM)-based methods mostly formulate it as direct...
By Xuanle Zhao, Xinyuan Cai, Xiang Cheng, Bo Xu
arXiv:2605.29888v2 Announce Type: replace-cross
Abstract: Reinforcement learning (RL) post-training has shown to improve reasoning in large language models (LLMs). However, there has been little expl...
By Minju Gwak, Minseo Kwak, Dongseok Lee, Guijin Son, Alan Ritter, Jaehyung Kim
arXiv:2512.03244v2 Announce Type: replace-cross
Abstract: Training process reward models (PRMs) requires step-level correctness labels, obtained either through expensive human annotation or by relyin...
By Salman Rahman, Sruthi Gorantla, Arpit Gupta, Swastik Roy, Nanyun Peng, Yang Liu
arXiv:2608.31167v1 Announce Type: cross
Abstract: Bridging model-based control and learned policies in long-horizon manipulation has harbored a silent disagreement: control executes specified objecti...
By Weiqi Wang, Zhi Li, Yudong Lei, David Martinez, Xiaofeng Gao, Yuxin Jiang, Chenfanfu Jiang, Yingnian Wu, Demetri Terzopoulos, Ran Gong
arXiv:2608.28859v1 Announce Type: cross
Abstract: Reasoning models do not stop when they know the answer. On DeepSeek-R1-Distill-Qwen-7B the chain of thought runs about twice as long as the model's o...
By Dylan Jayabahu, Tinuade Adeleke
The paper introduces FAB, an attack that uses meta‑learning to embed dormant adversarial behaviors into large language models (LLMs). These behaviors remain inactive until the model is finetuned by downstream users, at which point the model can exhibit unwanted actions such as unsolicited advertising, jailbreakability, or over‑refusal. FAB is shown to be effective across multiple LLMs and resilient to various finetuning settings.
By Thibaud Gloaguen, Mark Vero, Robin Staab, Martin Vechev
The paper introduces Common-Trace Factorized Constrained Proximal Policy Optimization (CT‑PPO) for optimizing online sensing‑session consolidation in multi‑tenant integrated sensing and communication (ISAC) networks. By formulating the problem as a constrained Markov decision process, CT‑PPO shares workload traces across stochastic policy replicas and uses leave‑one‑out discounted Monte Carlo returns to credit applicable actor factors, achieving higher macro returns and lower sensing‑resource costs than baseline methods. Ablation studies show that the common‑trace reward credit is the key contributor to performance gains, and the algorithm retains advantages across varying arrival loads without increasing deployment latency.
By Dang-Dung Vu
TEMPO is a unified model that adds temporally‑grounded capabilities to large audio‑language models, enabling timestamping of events, speakers, and sounds in audio, speech, and music. It introduces a supervised fine‑tuning stage featuring atomic timestamp tokens, a time‑aware projector with sinusoidal encodings, and a distance‑aware Gaussian loss, trained via a synthetic‑to‑real curriculum. Additionally, TEMPO employs reinforcement learning (GRPO) as a refinement step, and achieves state‑of‑the‑art performance on a benchmark of 10K samples across five timestamping tasks, surpassing Audio Flamingo Next and Qwen3‑Omni.
By Apoorva Kulkarni, Kaousheik Jayakumar, Sreyan Ghosh, Utathya Aich, Ramani Duraiswami, Dinesh Manocha