exploration strategies
Exploration strategies in reinforcement learning specify methods for exploring new actions to balance the trade-off between gathering information and maximizing rewards, essential for effective learning.
- Accelerating RL for LLM Reasoning with Optimal Advantage Regression
- Curious Causality-Seeking Agents Learn Meta Causal World
- GUI Exploration Lab: Enhancing Screen Navigation in Agents via Multi-Turn Reinforcement Learning
- Meta-learning how to Share Credit among Macro-Actions
- NoisyGRPO: Incentivizing Multimodal CoT Reasoning via Noise Injection and Bayesian Estimation