English(EN)Dual-Evolving Reinforcement Learning for Automated Reward Design
新的强化学习技术提升LLM推理和训练效率 · 跟踪7个来源
作者PulseAugur 编辑部·[7 个来源]·
几篇研究论文探讨了用于大型语言模型（LLM）的强化学习（RL）的进展，重点是提高推理能力和训练效率。一篇论文介绍了“双回滚强化学习”，以创建更鲁棒的奖励估计，从而提高Qwen和Nemotron等模型在数学和编码任务上的准确性。另一项研究提出了“动态Jensen-Shannon回放”（DyJR），以保持策略的可塑性并防止过早地过度承诺于单一解决方案，从而在各种基准测试中提高性能。此外，研究还探讨了动量在RLVR中的作用，表明特定的动量系数可以加速优化，并引入了“上下文价值感知上下文强化学习”（CV-ICRL）来增强测试时适应的稳定性。
AI
arXiv:2610.16593v1 Announce Type: new Abstract: Reinforcement learning with verifiable rewards (RLVR) implemented with Group Relative Policy Optimization (GRPO) has driven substantial progress in training reasoning language models. However, verifiable outcome rewards provide an i…
arXiv cs.AI
TIER_1English(EN)·Yuante Li, Junwei Quan, Kai Qiu, Kate Zhang, Yicheng Tao·
arXiv:2603.16157v2 Announce Type: replace-cross Abstract: While Reinforcement Learning (RL) enhances Large Language Model reasoning, on-policy algorithms like GRPO are sample-inefficient as they discard past rollouts. Existing experience replay methods address this by reusing acc…
arXiv:2512.13399v3 Announce Type: replace-cross Abstract: Crafting effective reward signals remains a central challenge in reinforcement learning (RL) for complex reasoning. Existing methods for automated reward optimization search with a fixed proposer which never learns from it…
arXiv:2610.19783v1 Announce Type: new Abstract: Reinforcement learning (RL) holds the promise of learning a policy that can outperform what is shown in the data. However, in practice, such systems suffer from instability and hyperparameter sensitivity. In contrast, supervised lea…
arXiv cs.LG
TIER_1English(EN)·Kaifei Wang, Zikang Shan, Liwei Wang, Yinyu Ye, Han Zhong·
arXiv:2610.20413v1 Announce Type: new Abstract: Momentum is a classical tool for accelerating convex optimization and a standard component of modern large language model training. Yet recent empirical work has questioned the usefulness of first-moment momentum in reinforcement le…
arXiv:2509.23209v2 Announce Type: replace Abstract: In-context reinforcement learning (ICRL) has emerged as an effective paradigm for test time adaptation to unseen tasks without parameter updates. However, existing ICRL methods can exhibit brittle and unstable adaptions, and the…
