中文
相关论文

相关论文: When Policy Entropy Constraint Fails: Preserving D…

200 篇论文

Reinforcement learning (RL) is a key paradigm for post-training large language models (LLMs), but the widely used Group Relative Policy Optimization (GRPO) often suffers from entropy collapse: exploration quickly disappears, policies…

机器学习 · 计算机科学 2026-05-19 Chen Wang , Zhaochun Li , Jionghao Bai , Hexuan Deng , Ge Lan , Yue Wang

This paper aims to overcome a major obstacle in scaling RL for reasoning with LLMs, namely the collapse of policy entropy. Such phenomenon is consistently observed across vast RL runs without entropy intervention, where the policy entropy…

Reinforcement learning with verifiable rewards (RLVR) has become an effective paradigm for improving the reasoning ability of large language models. However, widely used RLVR algorithms, such as GRPO, often suffer from entropy collapse,…

机器学习 · 计算机科学 2026-05-13 Huimin Xu , Shuai Zhao , Xiaobao Wu , Anh Tuan Luu

Policy entropy has emerged as a fundamental measure for understanding and controlling exploration in reinforcement learning with verifiable rewards (RLVR) for LLMs. However, existing entropy-aware methods mainly regulate entropy through…

Proximal Policy Optimization (PPO) has been positioned by recent literature as the canonical method for the RL part of Reinforcement Learning from Human Feedback (RLHF). PPO performs well empirically but has a heuristic motivation and…

机器学习 · 计算机科学 2026-02-10 Dipan Maity

Reinforcement Learning with Verifiable Rewards (RLVR) serves as a cornerstone technique for enhancing the reasoning capabilities of Large Language Models (LLMs). However, its training is often plagued by \emph{entropy collapse}, a rapid…

机器学习 · 计算机科学 2026-04-30 Zhezheng Hao , Hong Wang , Haoyang Liu , Jian Luo , Jiarui Yu , Hande Dong , Qiang Lin , Can Wang , Jiawei Chen

Policy gradient algorithms have driven many recent advancements in language model reasoning. An appealing property is their ability to learn from exploration on their own trajectories, a process crucial for fostering diverse and creative…

Reinforcement Learning with Verifiable Rewards (RLVR) has emerged as a critical method for enhancing the reasoning capabilities of Large Language Models (LLMs). However, continuous training often leads to policy entropy collapse,…

机器学习 · 计算机科学 2026-05-11 Kun Chen , Peng Shi , Fanfan Liu , Haibo Qiu , Zhixiong Zeng , Siqi Yang , Wenji Mao

While direct policy optimization methods exist, pioneering LLMs are fine-tuned with reinforcement learning from human feedback (RLHF) to generate better responses under the supervision of a reward model learned from preference data. One…

机器学习 · 计算机科学 2025-06-10 Chuheng Zhang , Wei Shen , Li Zhao , Xuyun Zhang , Xiaolong Xu , Wanchun Dou , Jiang Bian

Continual reinforcement learning must balance retention with adaptation, yet many methods still rely on \emph{single-model preservation}, committing to one evolving policy as the main reusable solution across tasks. Even when a previously…

机器学习 · 计算机科学 2026-04-20 Lute Lillo , Nick Cheney

Reinforcement Learning with Verifiable Rewards (RLVR) has proven to be a highly effective strategy for endowing Large Language Models (LLMs) with robust multi-step reasoning abilities. However, its design and optimizations remain tailored…

Reinforcement Learning with Verifiable Rewards (RLVR) has emerged as an important paradigm for unlocking reasoning capabilities in large language models, exemplified by the success of OpenAI o1 and DeepSeek-R1. Currently, Group Relative…

机器学习 · 计算机科学 2026-01-08 Shijie Zhang , Kevin Zhang , Zheyuan Gu , Xiang Guo , Rujun Guo , Shaoyu Liu , Guanjun Jiang , Xiaozhao Wang

Reinforcement learning with verifiable rewards (RLVR), particularly Group Relative Policy Optimization (GRPO), has advanced LLM reasoning. However, GRPO suffers from three credit assignment failures: uniform token-level granularity that…

机器学习 · 计算机科学 2026-05-07 Song Yu , Li Li , Wenwen Zhao , Zhisheng Yang

Reasoning ability has become a defining capability of Large Language Models (LLMs), with Reinforcement Learning with Verifiable Rewards (RLVR) emerging as a key paradigm to enhance it. However, RLVR training often suffers from policy…

机器学习 · 计算机科学 2026-04-20 Xiaoyun Zhang , Xiaojian Yuan , Di Huang , Wang You , Chen Hu , Jingqing Ruan , Ai Jian , Kejiang Chen , Xing Hu

Recently, reinforcement learning with verifiable rewards (RLVR) has been widely used for enhancing the reasoning abilities of large language models (LLMs). A core challenge in RLVR involves managing the exchange between entropy and…

计算与语言 · 计算机科学 2025-08-05 Jia Deng , Jie Chen , Zhipeng Chen , Wayne Xin Zhao , Ji-Rong Wen

Reinforcement learning with verifiable rewards (RLVR) has demonstrated superior performance in enhancing the reasoning capability of large language models (LLMs). However, this accuracy-oriented learning paradigm often suffers from entropy…

人工智能 · 计算机科学 2026-01-19 Hongye Cao , Zhixin Bai , Ziyue Peng , Boyan Wang , Tianpei Yang , Jing Huo , Yuyao Zhang , Yang Gao

RL training of multi-turn LLM agents is inherently unstable, and reasoning quality directly determines task performance. Entropy is widely used to track reasoning stability. However, entropy only measures diversity within the same input,…

Reinforcement-learning-based post-training has become a key approach for improving the reasoning ability of large language models, but its token-level learning signals remain poorly understood. This work studies their heterogeneity through…

计算与语言 · 计算机科学 2026-05-11 Gengyang Li , Zheng-Fan Wu , Siqi Bao , Yunfang Wu

Extending LLM context windows is hindered by scarce high-quality long-context data. Recent methods synthesize data with genuine long-range dependencies via information-theoretic verification, selecting contexts that reduce a base model's…

机器学习 · 计算机科学 2026-04-10 Junlong Jia , Ziyang Chen , Xing Wu , Chaochen Gao , TingHao Yu , Feng Zhang , Songlin Hu

Entropy serves as a critical metric for measuring the diversity of outputs generated by large language models (LLMs), providing valuable insights into their exploration capabilities. While recent studies increasingly focus on monitoring and…

机器学习 · 计算机科学 2026-02-04 Shumin Wang , Yuexiang Xie , Wenhao Zhang , Yuchang Sun , Yanxi Chen , Yaliang Li , Yanyong Zhang
‹ 上一页 1 2 3 10 下一页 ›