English
Related papers

Related papers: An operator view of policy gradient methods

200 papers

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

We study reinforcement learning in hybrid discrete-continuous action spaces, such as settings where the discrete component selects a regime (or index) and the continuous component optimizes within it -- a structure common in robotics,…

Machine Learning · Computer Science 2026-05-15 Matias Alvo , Daniel Russo , Yash Kanoria

We introduce a method for policy improvement that interpolates between the greedy approach of value-based reinforcement learning (RL) and the full planning approach typical of model-based RL. The new method builds on the concept of a…

Machine Learning · Statistics 2022-06-20 Shantanu Thakoor , Mark Rowland , Diana Borsa , Will Dabney , Rémi Munos , André Barreto

Policy gradient (PG) estimation becomes a challenge when we are not allowed to sample with the target policy but only have access to a dataset generated by some unknown behavior policy. Conventional methods for off-policy PG estimation…

Machine Learning · Statistics 2022-06-22 Chengzhuo Ni , Ruiqi Zhang , Xiang Ji , Xuezhou Zhang , Mengdi Wang

Existing value function approximation methods have been successfully used in many applications, but they often lack useful a priori error bounds. We propose a new approximate bilinear programming formulation of value function approximation,…

Artificial Intelligence · Computer Science 2010-06-15 Marek Petrik , Shlomo Zilberstein

Modern policy gradient algorithms such as Proximal Policy Optimization (PPO) rely on an arsenal of heuristics, including loss clipping and gradient clipping, to ensure successful learning. These heuristics are reminiscent of techniques from…

This paper presents a data-driven method to find a closed-loop optimal controller, which minimizes a specified infinite-horizon cost function for systems with unknown dynamics. Suppose the closed-loop optimal controller can be parameterized…

Machine Learning · Computer Science 2025-11-20 Wenjian Hao , Paulo C. Heredia , Shaoshuai Mou

Policy evaluation is an important instrument for the comparison of different algorithms in Reinforcement Learning (RL). However, even a precise knowledge of the value function $V^{\pi}$ corresponding to a policy $\pi$ does not provide…

Machine Learning · Computer Science 2026-01-21 Denis Belomestny , Ilya Levin , Alexey Naumov , Sergey Samsonov

We introduce Phasic Policy Gradient (PPG), a reinforcement learning framework which modifies traditional on-policy actor-critic methods by separating policy and value function training into distinct phases. In prior methods, one must choose…

Machine Learning · Computer Science 2020-09-10 Karl Cobbe , Jacob Hilton , Oleg Klimov , John Schulman

Both the optimal value function and the optimal policy can be used to model an optimal controller based on the duality established by the Bellman equation. Even with this duality, no parametric model has been able to output both policy and…

Systems and Control · Electrical Eng. & Systems 2020-06-02 Jicheng Shi , Yingzhao Lian , Colin N. Jones

Gradient boosting is a state-of-the-art prediction technique that sequentially produces a model in the form of linear combinations of simple predictors---typically decision trees---by solving an infinite-dimensional convex optimization…

Statistics Theory · Mathematics 2017-07-18 Gérard Biau , Benoît Cadre

We present the Generalized Policy Gradient (GPG) Theorem, specifically designed for Transformer-based policies. Notably, we demonstrate that both standard Policy Gradient Theorem and GRPO emerge as special cases within our GPG framework.…

Machine Learning · Computer Science 2025-12-12 Hangyu Mao , Guangting Dong , Zhicheng Dou

A common challenge in real-time operations is deciding whether to re-solve an optimization problem or continue using an existing solution. While modern data platforms may collect information at high frequencies, many real-time operations…

Machine Learning · Computer Science 2025-09-30 Rui Ai , Hugo De Oliveira Barbalho , Sirui Li , Alexei Robsky , David Simchi-Levi , Ishai Menache

Learned image reconstruction has become a pillar in computational imaging and inverse problems. Among the most successful approaches are learned iterative networks, which are formulated by unrolling classical iterative optimisation…

Image and Video Processing · Electrical Eng. & Systems 2025-12-10 Andreas Hauptmann , Ozan Öktem

We introduce a new algorithm for multi-objective reinforcement learning (MORL) with linear preferences, with the goal of enabling few-shot adaptation to new tasks. In MORL, the aim is to learn policies over multiple competing objectives…

Machine Learning · Computer Science 2019-11-07 Runzhe Yang , Xingyuan Sun , Karthik Narasimhan

We study the problem of programmatic reinforcement learning, in which policies are represented as short programs in a symbolic language. Programmatic policies can be more interpretable, generalizable, and amenable to formal verification…

Machine Learning · Computer Science 2021-01-21 Abhinav Verma , Hoang M. Le , Yisong Yue , Swarat Chaudhuri

We study how the behavior of deep policy gradient algorithms reflects the conceptual framework motivating their development. To this end, we propose a fine-grained analysis of state-of-the-art methods based on key elements of this…

Machine Learning · Computer Science 2020-05-26 Andrew Ilyas , Logan Engstrom , Shibani Santurkar , Dimitris Tsipras , Firdaus Janoos , Larry Rudolph , Aleksander Madry

Recently, foundation models such as OpenAI's O1 and O3, along with DeepSeek's R1, have demonstrated strong reasoning capacities and problem-solving skills acquired through large-scale reinforcement learning (RL), with wide applications in…

Machine Learning · Computer Science 2025-04-29 Tianbing Xu

Very recently proximal policy optimization (PPO) algorithms have been proposed as first-order optimization methods for effective reinforcement learning. While PPO is inspired by the same learning theory that justifies trust region policy…

Machine Learning · Computer Science 2018-04-20 Gang Chen , Yiming Peng , Mengjie Zhang

We revisit Group Relative Policy Optimization (GRPO) in both on-policy and off-policy optimization regimes. Our motivation comes from recent work on off-policy Proximal Policy Optimization (PPO), which improves training stability, sampling…