Traditional Reinforcement Learning

Traditional reinforcement learning (RL) aims to train agents to make optimal decisions in dynamic environments by maximizing cumulative rewards. Current research focuses on improving sample efficiency and addressing limitations like the need for manually designed reward functions, exploring techniques such as inverse reinforcement learning (IRL), generative adversarial imitation learning (GAIL), and the use of large language models (LLMs) for policy learning and control. These advancements are impacting diverse fields, from robotics and game playing to optimizing complex systems like satellite networks and drug design, by enabling more efficient and adaptable agent training. Furthermore, research is actively exploring methods to handle multi-objective scenarios and mitigate challenges like catastrophic forgetting in continual learning settings.

Papers

February 1, 2024

Dense Reward for Free in Reinforcement Learning from Human Feedback
Alex J. Chan, Hao Sun, Samuel Holt, Mihaela van der Schaar
Reinforcement Learning Human Feedback Reward Model Reward Shaping Traditional Reinforcement Learning Dense Reward

January 22, 2024

P2DT: Mitigating Forgetting in task-incremental Learning with progressive prompt Decision Transformer
Zhiyuan Wang, Xiaoyang Qu, Jing Xiao, Bokui Chen, Jianzong Wang
Catastrophic Forgetting Faulty Negative Mitigation Decision Transformer Traditional Reinforcement Learning Task Specific Policy Task Incremental Learning

July 18, 2023

REX: Rapid Exploration and eXploitation for AI Agents
Rithesh Murthy, Shelby Heinecke, Juan Carlos Niebles, Zhiwei Liu, Le Xue, Weiran Yao, Yihao Feng, Zeyuan Chen, Akash Gokul, Devansh Arpit, Ran Xu, Phil Mui, Huan Wang, Caiming Xiong, Silvio Savarese
AI Agent Real World eXploitation Traditional Reinforcement Learning Exponential Mechanism Rapid Exploration

March 20, 2023

Reflexion: Language Agents with Verbal Reinforcement Learning
Noah Shinn, Federico Cassano, Edward Berman, Ashwin Gopinath, Karthik Narasimhan, Shunyu Yao
Language Agent Goal Oriented Traditional Reinforcement Learning Natural Language Reasoning Language Feedback

March 6, 2023

Perspectives on the Social Impacts of Reinforcement Learning with Human Feedback
Gabrielle Kaili-May Liu
Reinforcement Learning Artificial Intelligence Human Feedback Deep Network Synthesized View Traditional Reinforcement Learning Societal Impact

March 5, 2023

Local Environment Poisoning Attacks on Federated Reinforcement Learning
Evelyn Ma, Praneet Rathi, S. Rasoul Etesami
Reinforcement Learning Poisoning Attack Federated Reinforcement Learning Traditional Reinforcement Learning

February 27, 2023

Reinforcement Learning with Depreciating Assets
Taylor Dohmen, Ashutosh Trivedi
Reinforcement Learning Optimal Policy Traditional Reinforcement Learning Portfolio Management Exponential Discounting

January 31, 2023

Enabling surrogate-assisted evolutionary reinforcement learning via policy embedding
Lan Tang, Xiaxi Li, Jinyuan Zhang, Guiying Li, Peng Yang, Ke Tang
Reinforcement Learning Evolutionary Algorithm Health Policy Recommendation Traditional Reinforcement Learning State of the Art Algorithm

January 14, 2023

Risk-Averse Reinforcement Learning via Dynamic Time-Consistent Risk Measures
Xian Yu, Siqian Shen
Traditional Reinforcement Learning Infinite Horizon Risk Measure Risk Averse Reinforcement Learning

December 15, 2022

Residual Policy Learning for Powertrain Control
Lindsey Kerbel, Beshah Ayalew, Andrej Ivanco, Keith Loiselle
Traditional Reinforcement Learning Residual Reinforcement Learning Residual Policy

October 23, 2022

Active Predictive Coding: A Unified Neural Framework for Learning Hierarchical World Models for Perception and Planning
Rajesh P. N. Rao, Dimitrios C. Gklezakos, Vishwas Sathish
Reinforcement Learning Task Planning World Model Perception Aware Predictive Coding Traditional Reinforcement Learning Compositional Representation Unified Neural

October 14, 2022

Geometric Reinforcement Learning For Robotic Manipulation
Naseem Alhousani, Matteo Saveriano, Ibrahim Sevinc, Talha Abdulkuddus, Hatice Kose, Fares J. Abu-Dakka
Reinforcement Learning Robotic Manipulation Policy Iteration Traditional Reinforcement Learning Riemannian Geometry Manipulation Skill

October 12, 2022

Contrastive Retrospection: honing in on critical steps for rapid learning and generalization in RL
Chen Sun, Wannan Yang, Thomas Jiralerspong, Dane Malenfant, Benjamin Alsbury-Nealy, Yoshua Bengio, Blake Richards
Contrastive Learning Strong Generalization Contrastive Loss Intrinsic Reward Fast Learning Traditional Reinforcement Learning Critical Scenario

September 27, 2022

DCE: Offline Reinforcement Learning With Double Conservative Estimates
Chen Zhao, Kai Xing Huang, Chun yuan
Offline Reinforcement Learning Traditional Reinforcement Learning Conservative Value Estimation

September 17, 2022

Sub-optimal Policy Aided Multi-Agent Reinforcement Learning for Flocking Control
Yunbo Qiu, Yue Jin, Jian Wang, Xudong Zhang
Multi Agent Reinforcement Learning Traditional Reinforcement Learning Flocking Control Sub Optimal Policy

August 4, 2022

Backward Imitation and Forward Reinforcement Learning via Bi-directional Model Rollouts
Yuxin Pan, Fangzhen Lin
Reinforcement Learning Traditional Reinforcement Learning State of the Art Imitation Model Rollouts

July 13, 2022

Brick Tic-Tac-Toe: Exploring the Generalizability of AlphaZero to Novel Test Environments
John Tan Chong Min, Mehul Motani
Reinforcement Learning Stronger Generalizability Traditional Reinforcement Learning AlphaZero Based Artificial Intelligence Test Environment Tic Tac Toe

June 16, 2022

How to talk so AI will learn: Instructions, descriptions, and autonomy
Theodore R Sumers, Robert D Hawkins, Mark K Ho, Thomas L Griffiths, Dylan Hadfield-Menell
Artificial Intelligence Reward Function Human Instruction Legal Autonomy Description Library Traditional Reinforcement Learning Social Learning Contextual Bandit Setting Behavioural Experiment

January 15, 2022

Recursive Least Squares Advantage Actor-Critic Algorithms
Yuan Wang, Chunyuan Zhang, Tianzong Yu, Meng Ma
Deep Reinforcement Learning Least Square Actor Critic Algorithm Recursive Algorithm Traditional Reinforcement Learning Advantage Actor Critic Apical 2 Chamber

November 8, 2021

Dueling RL: Reinforcement Learning with Trajectory Preferences
Aldo Pacchiano, Aadirupa Saha, Jonathan Lee
Reinforcement Learning Regret Guarantee Traditional Reinforcement Learning Human Like RL Non Markovian Reward Trajectory Preference