English
Related papers

Related papers: On Proximal Policy Optimization's Heavy-tailed Gra…

200 papers

Preference optimization has become a central paradigm for aligning large language models with human feedback. Direct Preference Optimization (DPO) simplifies reinforcement learning from human feedback by directly optimizing pairwise…

Machine Learning · Computer Science 2026-05-05 Inoussa Mouiche

We propose policy gradient algorithms for solving a risk-sensitive reinforcement learning (RL) problem in on-policy as well as off-policy settings. We consider episodic Markov decision processes, and model the risk using the broad class of…

Machine Learning · Computer Science 2024-06-25 Nithia Vijayan , Prashanth L. A

This paper considers the problem of learning safe policies in the context of reinforcement learning (RL). In particular, we consider the notion of probabilistic safety. This is, we aim to design policies that maintain the state of the…

Machine Learning · Computer Science 2023-04-20 Weiqin Chen , Dharmashankar Subramanian , Santiago Paternain

Classical reinforcement learning (RL) aims to optimize the expected cumulative rewards. In this work, we consider the RL setting where the goal is to optimize the quantile of the cumulative rewards. We parameterize the policy controlling…

Machine Learning · Computer Science 2022-02-17 Jinyang Jiang , Jiaqiao Hu , Yijie Peng

Reinforcement learning (RL) is pivotal for enabling large language models (LLMs) to generate long chains of thought (CoT) for complex tasks like math and reasoning. However, Proximal Policy Optimization (PPO), effective in many RL…

Machine Learning · Computer Science 2025-03-04 Yufeng Yuan , Yu Yue , Ruofei Zhu , Tiantian Fan , Lin Yan

This paper studies the distributed optimization problem under the influence of heavy-tailed gradient noises. Here, a heavy-tailed noise means that the noise does not necessarily satisfy the bounded variance assumption. Instead, it satisfies…

Optimization and Control · Mathematics 2025-05-12 Chao Sun , Huiming Zhang , Bo Chen , Li Yu

Existing policy-gradient methods for auto-regressive language models typically select subsequent tokens one at a time as actions in the policy. While effective for many generation tasks, such an approach may not fully capture the structure…

Computation and Language · Computer Science 2026-02-17 Mufan Xu , Kehai Chen , Xuefeng Bai , Zhengyu Niu , Muyun Yang , Tiejun Zhao , Min Zhang

Differentially Private Stochastic Gradient Descent (DPSGD) is widely utilized to preserve training data privacy in deep learning, which first clips the gradients to a predefined norm and then injects calibrated noise into the training…

Machine Learning · Computer Science 2024-05-29 Haichao Sha , Yang Cao , Yong Liu , Yuncheng Wu , Ruixuan Liu , Hong Chen

Despite Proximal Policy Optimization (PPO) dominating policy gradient methods -- from robotic control to game AI -- its static trust region forces a brittle trade-off: aggressive clipping stifles early exploration, while late-stage updates…

Machine Learning · Computer Science 2025-05-26 Ben Rahman

We consider the problem of learning control policies that optimize a reward function while satisfying constraints due to considerations of safety, fairness, or other costs. We propose a new algorithm, Projection-Based Constrained Policy…

Machine Learning · Computer Science 2020-10-08 Tsung-Yen Yang , Justinian Rosca , Karthik Narasimhan , Peter J. Ramadge

Group Relative Policy Optimization (GRPO) has emerged as a promising critic-free reinforcement learning paradigm for reasoning tasks. However, standard GRPO employs a coarse-grained credit assignment mechanism that propagates group-level…

Computation and Language · Computer Science 2026-01-13 Ziheng Li , Liu Kang , Feng Xiao , Luxi Xing , Qingyi Si , Zhuoran Li , Weikang Gong , Deqing Yang , Yanghua Xiao , Hongcheng Guo

Group Relative Policy Optimization (GRPO) has significantly enhanced the reasoning capability of large language models by optimizing the arithmetic mean of token-level rewards. Unfortunately, GRPO is observed to suffer from unstable policy…

Computation and Language · Computer Science 2025-10-21 Yuzhong Zhao , Yue Liu , Junpeng Liu , Jingye Chen , Xun Wu , Yaru Hao , Tengchao Lv , Shaohan Huang , Lei Cui , Qixiang Ye , Fang Wan , Furu Wei

The role of reinforcement learning (RL) in enhancing the reasoning of large language models (LLMs) is becoming increasingly significant. Despite the success of RL in many scenarios, there are still many challenges in improving the reasoning…

Artificial Intelligence · Computer Science 2024-12-25 Jiacai Liu , Chaojie Wang , Chris Yuhao Liu , Liang Zeng , Rui Yan , Yiwen Sun , Yang Liu , Yahui Zhou

Policy gradient methods have enabled deep reinforcement learning (RL) to approach challenging continuous control problems, even when the underlying systems involve highly nonlinear dynamics that generate complex non-smooth optimization…

Machine Learning · Computer Science 2024-05-29 Tao Wang , Sylvia Herbert , Sicun Gao

Policy Gradient (PG) algorithms are among the best candidates for the much-anticipated applications of reinforcement learning to real-world control tasks, such as robotics. However, the trial-and-error nature of these methods poses safety…

Machine Learning · Computer Science 2022-06-20 Matteo Papini , Matteo Pirotta , Marcello Restelli

The policy represented by the deep neural network can overfit the spurious features in observations, which hamper a reinforcement learning agent from learning effective policy. This issue becomes severe in high-dimensional state, where the…

Machine Learning · Computer Science 2023-05-01 Md Masudur Rahman , Yexiang Xue

Direct Preference Optimization (DPO) has been proposed as a promising alternative to Proximal Policy Optimization (PPO) based Reinforcement Learning with Human Feedback (RLHF). However, empirical evaluations consistently reveal suboptimal…

Machine Learning · Computer Science 2025-03-03 Qinwei Ma , Jingzhe Shi , Can Jin , Jenq-Neng Hwang , Serge Belongie , Lei Li

Latent reasoning offers a more efficient alternative to explicit reasoning by compressing intermediate reasoning into continuous representations and substantially shortening reasoning chains. However, existing latent reasoning methods…

Machine Learning · Computer Science 2026-05-01 Jingcheng Deng , Zihao Wei , Liang Pang , Junhong Wu , Shicheng Xu , Zenghao Duan , Huawei Shen

We propose policy gradient algorithms which learn risk-sensitive policies in a reinforcement learning (RL) framework. Our proposed algorithms maximize the distortion risk measure (DRM) of the cumulative reward in an episodic Markov decision…

Machine Learning · Computer Science 2024-02-06 Nithia Vijayan , Prashanth L. A

Reinforcement learning with verifiable rewards (RLVR) plays a pivotal role in improving the reasoning ability of large language models. However, widely used PPO surrogate objectives are fundamentally local, as they rely on a local…

Machine Learning · Computer Science 2026-05-21 Deokgyu Yoon , Hyungkyu Kang , Joongkyu Lee , Byeongchan Kim , Gyungin Shin , Sungrae Park , Min-hwan Oh