Similar Links:

• Go-Explore: First return, then explore

• First return, then explore

• Go-Explore: a New Approach for Hard-Exploration Problems

• Never Give Up: Learning Directed Exploration Strategies

• Observe and Look Further: Achieving Consistent Performance on Atari

• Unifying Count-Based Exploration and Intrinsic Motivation

• Exploration by Random Network Distillation

• Benchmarking Bonus-Based Exploration Methods on the Arcade Learning Environment

• E3B: Exploration via Elliptical Episodic Bonuses

• General Exploratory Bonus for Optimistic Exploration in RLHF

• Semantic Exploration from Language Abstractions and Pretrained Representations

• Meta-RL Induces Exploration in Language Agents

• Meta Reinforcement Learning

• Exploration Strategies in Deep Reinforcement Learning

• Planning to Explore via Self-Supervised World Models

• Deep Exploration via Bootstrapped DQN

• Replay-Guided Adversarial Environment Design

• Asymmetric self-play for automatic goal discovery in robotic manipulation

• Curiosity in hindsight

• AIs May Not Need Sleep or Play

• Search: GS; Google; site