Language Model Self-improvement by Reinforcement Learning Contemplation

Jing-Cheng Pang, Pengyuan Wang, Kaiyuan Li +4

Large Language Models (LLMs) have exhibited remarkable performance across various natural language processing (NLP) tasks. However, fine-tuning these models often necessitates subs…