Awesome-AgenticLLM-RL-Papers

The repository accompanies a peer-reviewed survey paper examining how reinforcement learning enables LLMs to act as agents. It focuses on algorithmic approaches that improve decision-making, alignment, and reasoning through reward signals and policy optimization.

Content is structured around major method families, highlighting mechanisms such as policy ratio clipping, KL penalties, and group-based rewards. Each entry links to original papers and available code or models for further exploration.

Intended for machine learning researchers and practitioners who need an organized reference when studying or implementing agentic RL strategies in language models.