Policy Based Reinforcement Learning

Policy-based reinforcement learning (RL) focuses on directly learning optimal action policies for agents interacting with an environment, aiming to maximize cumulative rewards. Current research emphasizes improving the efficiency and robustness of these methods, particularly in complex settings like multi-turn language interactions with large language models and program synthesis, often employing actor-critic architectures and exploring value-based alternatives. This work is significant because it addresses challenges in sample efficiency, uncertainty quantification, and explainability, leading to more reliable and trustworthy RL agents with applications ranging from human-computer interaction to robust control systems.

Papers

October 23, 2024

Entity-based Reinforcement Learning for Autonomous Cyber Defence
Isaac Symes Thompson, Alberto Caron, Chris Hicks, Vasilios Mavroudis
Deep Reinforcement Learning Autonomous Cyber Transformer Based Policy Policy Based Reinforcement Learning Autonomous Network

November 30, 2023

LMRL Gym: Benchmarks for Multi-Turn Reinforcement Learning with Language Models
Marwa Abdulhai, Isadora White, Charlie Snell, Charles Sun, Joey Hong, Yuexiang Zhai, Kelvin Xu, Sergey Levine
Language Model Reinforcement Learning New Benchmark Language Agent Goal Oriented Multi Turn Conversation MR iNet Gym Policy Based Reinforcement Learning

October 4, 2023

$\mathcal{B}$-Coder: Value-Based Deep Reinforcement Learning for Program Synthesis
Zishun Yu, Yunzhe Tao, Liyu Chen, Tao Sun, Hongxia Yang
Reinforcement Learning Program Synthesis Value Based Reinforcement Learning Value Based Deep Reinforcement Learning Value Based Policy Based Reinforcement Learning

March 14, 2023

Adaptive Policy Learning for Offline-to-Online Reinforcement Learning
Han Zheng, Xufang Luo, Pengfei Wei, Xuan Song, Dongsheng Li, Jing Jiang
Reinforcement Learning Offline Reinforcement Learning Offline Dataset Offline to Online Reinforcement Learning Policy Based Reinforcement Learning

October 7, 2022

How to Enable Uncertainty Estimation in Proximal Policy Optimization
Eugene Bykovets, Yannick Metz, Mennatallah El-Assady, Daniel A. Keim, Joachim M. Buhmann
Deep Reinforcement Learning Proximal Policy Optimization Actor Critic Reinforcement Learning Quantified Uncertainty Policy Based Reinforcement Learning Policy Uncertainty

July 18, 2022

Boolean Decision Rules for Reinforcement Learning Policy Summarisation
James McCarthy, Rahul Nair, Elizabeth Daly, Radu Marinescu, Ivana Dusparic
Reinforcement Learning DQN Agent Rule Based Model Policy Based Reinforcement Learning Boolean Rule

February 19, 2022

A Behavior Regularized Implicit Policy for Offline Reinforcement Learning
Shentao Yang, Zhendong Wang, Huangjie Zheng, Yihao Feng, Mingyuan Zhou
Offline Reinforcement Learning BEHAVIOR Explanation State Action Pair Policy Based Reinforcement Learning Implicit Policy

November 30, 2021

Model-Free $\mu$ Synthesis via Adversarial Reinforcement Learning
Darioush Keivan, Aaron Havens, Peter Seiler, Geir Dullerud, Bin Hu
Model Free Adversarial Reinforcement Learning Reactive Synthesis Policy Based Reinforcement Learning

November 5, 2021

An Algorithmic Theory of Metacognition in Minds and Machines
Rylan Schaeffer
Reinforcement Learning New Machine Human Mind Actor Critic Policy Based Reinforcement Learning