Intrinsic Exploration

Intrinsic exploration in reinforcement learning aims to enhance an agent's ability to discover rewarding states in environments with sparse or delayed feedback, improving learning efficiency. Current research focuses on developing novel intrinsic reward functions, often leveraging techniques like Bayesian surprise, chaotic dynamics, or generative adversarial networks, and integrating these with advanced reinforcement learning algorithms such as TD3 and various actor-critic methods. These advancements address limitations of existing methods, such as instability in optimizing non-stationary rewards or poor performance in stochastic environments, leading to more efficient and robust exploration strategies. Ultimately, improved intrinsic exploration promises to accelerate progress in complex reinforcement learning tasks and enable more effective transfer learning across diverse domains.

Papers

May 28, 2024

Individual Contributions as Intrinsic Exploration Scaffolds for Multi-agent Reinforcement Learning
Xinran Li, Zifan Liu, Shibo Chen, Jun Zhang
Multi Agent Reinforcement Learning Intrinsic Reward Client Contribution Sparse Reward Environment Intrinsic Exploration

May 15, 2024

Chaos-based reinforcement learning with TD3
Toshitaka Matsuki, Yusuke Sakemi, Kazuyuki Aihara
Reinforcement Learning Delayed Deep Deterministic Policy Gradient Intrinsic Exploration

October 27, 2023

Improving Intrinsic Exploration by Creating Stationary Objectives
Roger Creus Castanyer, Joshua Romoff, Glen Berseth
Active Exploration Exploration Problem Time Varying Objective Intrinsic Exploration

May 24, 2023

Successor-Predecessor Intrinsic Exploration
Changmin Yu, Neil Burgess, Maneesh Sahani, Samuel J. Gershman
Intrinsic Reward Intrinsic Exploration

February 8, 2023

Investigating the role of model-based learning in exploration and transfer
Jacob Walker, Eszter Vértes, Yazhe Li, Gabriel Dulac-Arnold, Ankesh Anand, Théophane Weber, Jessica B. Hamrick
Reinforcement Learning Transfer Learning Integral Role Environment Exploration Reward Signal Model Learning Model Based Agent Intrinsic Exploration

November 18, 2022

Curiosity in Hindsight: Intrinsic Exploration in Stochastic Environments
Daniel Jarrett, Corentin Tallec, Florent Altché, Thomas Mesnard, Rémi Munos, Michal Valko
Sparse Reward Intrinsic Reward Stochastic Environment Curiosity Inducing Situation Hindsight Instruction Reward Free Intrinsic Exploration Curiosity Driven Learning

October 1, 2022

Deep Intrinsically Motivated Exploration in Continuous Control
Baturay Saglam, Suleyman S. Kozat
Continuous Control Intrinsic Motivation Pure Exploration Reinforcement Learning Paradigm Intrinsic Exploration

June 28, 2022

GAN-based Intrinsic Exploration For Sample Efficient Reinforcement Learning
Doğay Kamar, Nazım Kemal Üre, Gözde Ünal
Reinforcement Learning Sparse Reward Efficient Exploration Sample Efficient Reinforcement Learning Adversarial Reward Intrinsic Exploration

June 23, 2022

CGAR: Critic Guided Action Redistribution in Reinforcement Leaning
Tairan Huang, Xu Li, Hao Li, Mingming Sun, Ping Li
Positive Reinforcement Policy Actor Critic Policy Distribution Intrinsic Exploration Understanding Reinforcement

February 17, 2022

Improving Intrinsic Exploration with Language Abstractions
Jesse Mu, Victor Zhong, Roberta Raileanu, Minqi Jiang, Noah Goodman, Tim Rocktäschel, Edward Grefenstette
Reinforcement Learning Intrinsic Reward Intrinsic Exploration Language Abstraction