Dynamic Bottleneck for Robust Self-Supervised Exploration

2021/10/20 by Chenjia Bai, Bai, Chenjia, Lingxiao Wang +11 · 2 citations
Computer Science · Decision Sciences · #Advanced Bandit Algorithms Research #Artificial Intelligence (cs.AI) #Artificial Intelligence in Games #FOS: Computer and information sciences #Machine Learning (cs.LG) #Reinforcement Learning in Robotics

paper · pdf · doi:10.48550/arxiv.2110.10735

openalex publication_date 2021/10/20 · openalex created_date 2025/10/10 · openalex updated_date 2026/07/28

Abstract

Exploration methods based on pseudo-count of transitions or curiosity of dynamics have achieved promising results in solving reinforcement learning with sparse rewards. However, such methods are usually sensitive to environmental dynamics-irrelevant information, e.g., white-noise. To handle such dynamics-irrelevant information, we propose a Dynamic Bottleneck (DB) model, which attains a dynamics-relevant representation based on the information-bottleneck principle. Based on the DB model, we further propose DB-bonus, which encourages the agent to explore state-action pairs with high information gain. We establish theoretical connections between the proposed DB-bonus, the upper confidence bound (UCB) for linear case, and the visiting count for tabular case. We evaluate the proposed method on Atari suits with dynamics-irrelevant noises. Our experiments show that exploration with DB bonus outperforms several state-of-the-art exploration methods in noisy environments.

Citations

• The information bottleneck method
• Opening the Black Box of Deep Neural Networks via Information
• Optimistic Exploration even with a Pessimistic Initialisation
• Improved Baselines with Momentum Contrastive Learning
• Momentum Contrast for Unsupervised Visual Representation Learning
• Provably Efficient Reinforcement Learning with Linear Function Approximation
• CURL: Contrastive Unsupervised Representations for Reinforcement Learning

Cited by

• Enhancing Q-Value Updates in Deep Q-Learning via Successor-State Prediction
• Wonder Wins Ways: Curiosity-Driven Exploration through Multi-Agent Contextual Calibration

Related

• Self-supervised Sequential Information Bottleneck for Robust Exploration in Deep Reinforcement Learning
• Planning with Exploration: Addressing Dynamics Bottleneck in Model-based Reinforcement Learning
• Context-Dependent Upper-Confidence Bounds for Directed Exploration
• Drop-Bottleneck: Learning Discrete Compressed Representation for Noise-Robust Exploration
• Learning to Explore in Diverse Reward Settings via Temporal-Difference-Error Maximization