Back to Basics: Revisiting REINFORCE Style Optimization for Learning from Human Feedback in LLMs

Arxiv Papers - Back to Basics: Revisiting REINFORCE Style Optimization for Learning from Human Feedback in LLMs