English
Related papers

Related papers: Rethinking Importance Sampling in LLM Policy Optim…

200 papers

Reinforcement Learning with Verifiable Rewards (RLVR) is commonly based on group sampling to estimate advantages and stabilize policy updates. In practice, computational limits often rule out very large groups, so training proceeds with…

Group Relative Policy Optimization (GRPO) has become the dominant method for reinforcement learning with verifiable rewards in large language models, but it suffers from two critical limitations: gradient vanishing and diversity collapse.…

Machine Learning · Computer Science 2026-05-20 Khiem Le , Phuc Nguyen , Youssef Mroueh , Chi-Heng Lin , Shangqian Gao , Ting Hua , Nitesh V. Chawla

Large language models (LLMs) demonstrate strong reasoning abilities via Chain-of-Thought (CoT), but their token-level generation encourages local decisions and lacks global planning, often leading to redundant or inaccurate reasoning.…

Recent large reasoning models (LRMs) driven by reinforcement learning algorithms (e.g., GRPO) have achieved remarkable performance on challenging reasoning tasks. However, these models suffer from overthinking, generating unnecessarily long…

Artificial Intelligence · Computer Science 2026-03-03 Gang Li , Yan Chen , Ming Lin , Tianbao Yang

Reinforcement Learning with Verifiable Rewards (RLVR) enhances Large Language Model (LLM) reasoning but suffers from advantage collapse on ``hard samples'' where all rollouts fail. This lack of variance eliminates crucial learning signals.…

Machine Learning · Computer Science 2026-05-08 Xinyu Lu , Kaiqi Zhang , Jinglin Yang , Boxi Cao , Yaojie Lu , Hongyu Lin , Min He , Xianpei Han , Le Sun

Group Relative Policy Optimization (GRPO) has been a key driver of recent progress in reinforcement learning with verifiable rewards (RLVR) for large language models, but it is typically trained in a low-staleness, near-on-policy regime…

Machine Learning · Computer Science 2026-05-19 Minghao Tian , Yunfei Xie , Chen Wei

Reinforcement learning (RL) has become popular in enhancing the reasoning capabilities of large language models (LLMs), with Group Relative Policy Optimization (GRPO) emerging as a widely used algorithm in recent systems. Despite GRPO's…

Machine Learning · Computer Science 2025-05-27 Wenlong Deng , Yi Ren , Muchen Li , Danica J. Sutherland , Xiaoxiao Li , Christos Thrampoulidis

Group Relative Policy Optimization(GRPO) has become a cornerstone of modern reinforcement learning alignment, prized for its efficacy in foregoing an explicit value-critic by leveraging reward normalization across sampled trajectory…

Computation and Language · Computer Science 2026-05-29 Redacted by arXiv

Group Relative Policy Optimization (GRPO), a prominent algorithm within the Reinforcement Learning from Verifiable Rewards (RLVR) framework, has achieved strong results in improving the reasoning capabilities of large language models…

Machine Learning · Computer Science 2026-05-21 Xixiang He , Qiyao Sun , Ao Cheng , Xingming Li , Xuanyu Ji , Hailun Lu , Runke Huang , Qingyong Hu

Reward-based alignment methods for large language models (LLMs) face two key limitations: vulnerability to reward hacking, where models exploit flaws in the reward signal; and reliance on brittle, labor-intensive prompt engineering when…

Computation and Language · Computer Science 2025-05-20 Zae Myung Kim , Chanwoo Park , Vipul Raheja , Suin Kim , Dongyeop Kang

Reinforcement learning (RL) for large language models (LLMs) is dominated by the cost of rollout generation, which has motivated the use of low-precision rollouts (e.g., FP8) paired with a BF16 trainer to improve throughput and reduce…

Machine Learning · Statistics 2026-05-15 Jiajun Zhou , Wei Shao , Lingchao Zheng , Yuwei Fan , Ngai Wong

Reinforcement learning with verifiable rewards (RLVR) has significantly advanced the reasoning ability of vision-language models (VLMs). However, the inherent text-dominated nature of VLMs often leads to insufficient visual faithfulness,…

Computer Vision and Pattern Recognition · Computer Science 2026-05-25 Zengbin Wang , Feng Xiong , Liang Lin , Xuecai Hu , Yong Wang , Yanlin Wang , Man Zhang , Xiangxiang Chu

We explore efficient estimation of statistical quantities, particularly rare event probabilities, for stochastic reaction networks. Consequently, we propose an importance sampling (IS) approach to improve the Monte Carlo (MC) estimator…

Numerical Analysis · Mathematics 2024-03-12 Chiheb Ben Hammouda , Nadhir Ben Rached , Raúl Tempone , Sophia Wiechert

Reinforcement learning methods such as GRPO have seen great popularity in LLM post-training. In GRPO, models produce completions to a set of prompts, which are rewarded, and the policy is updated towards the relatively high reward…

Machine Learning · Computer Science 2026-05-22 Nikolay Blagoev , Oğuzhan Ersoy , Wendelin Boehmer , Lydia Yiyu Chen

Contemporary reinforcement learning with verifiable reward methods post-train language models on multi-step reasoning by assigning a single outcome reward uniformly across all tokens in a trajectory. Such uniform assignment ignores which…

Artificial Intelligence · Computer Science 2026-05-27 Ankur Samanta , Akshayaa Magesh , Ayush Jain , Youliang Yu , Daniel Jiang , Kavosh Asadi , Kaveh Hassani , Paul Sajda , Jalaj Bhandari , Yonathan Efroni

Reinforcement Learning (RL) has significantly improved large language model reasoning, but existing RL fine-tuning methods rely heavily on heuristic techniques such as entropy regularization and reweighting to maintain stability. In…

Computation and Language · Computer Science 2026-05-26 Shiqi Liu , Zeyu He , Guojian Zhan , Letian Tao , Zhilong Zheng , Jiang Wu , Yinuo Wang , Yang Guan , Kehua Sheng , Bo Zhang , Keqiang Li , Jingliang Duan , Shengbo Eben Li

Training large language models (LLMs) with reinforcement learning (RL) methods such as PPO and GRPO commonly relies on ratio clipping to stabilise updates. While effective at preventing instability, clipping discards information, introduces…

Machine Learning · Computer Science 2026-02-02 Madeleine Dwyer , Adam Sobey , Adriane Chapman

Self-supervised reinforcement learning (RL) presents a promising approach for enhancing the reasoning capabilities of Large Language Models (LLMs) without reliance on expensive human-annotated data. However, we find that existing methods…

Artificial Intelligence · Computer Science 2025-12-16 Bizhe Bai , Hongming Wu , Peng Ye , Tao Chen

State-of-the-art large multi-modal models (LMMs) face challenges when processing high-resolution images, as these inputs are converted into enormous visual tokens, many of which are irrelevant to the downstream task. In this paper, we…

Computer Vision and Pattern Recognition · Computer Science 2026-04-21 Xinyu Huang , Yuhao Dong , Weiwei Tian , Bo Li , Rui Feng , Ziwei Liu

Extending large language models (LLMs) to low-resource languages often incurs an "alignment tax": improvements in the target language come at the cost of catastrophic forgetting in general capabilities. We argue that this trade-off arises…

Computation and Language · Computer Science 2026-05-15 Zeli Su , Ziyin Zhang , Zhou Liu , Xuexian Song , Zhankai Xu , Longfei Zheng , Xiaolu Zhang , Rong Fu , Guixian Xu , Wentao Zhang