PERL: Parameter Efficient Reinforcement Learning from Human Feedback
Paper
•
2403.10704
•
Published
•
59
papers
-
-
RLAIF: Scaling Reinforcement Learning from Human Feedback with AI Feedback
Paper • 2309.00267 • Published • 51 -
Absolute Zero: Reinforced Self-play Reasoning with Zero Data
Paper • 2505.03335 • Published • 186 -
Beyond the 80/20 Rule: High-Entropy Minority Tokens Drive Effective Reinforcement Learning for LLM Reasoning
Paper • 2506.01939 • Published • 185 -
Reinforcement Pre-Training
Paper • 2506.08007 • Published • 262