Necolizer/awesome-rl-for-agents

This list focuses on applying RL to train or tune Large Language Model (LLM) and Multimodal LLM (MLLM) agents, emphasizing research into computer-using, research-driven, and tool-integrated agent behaviors. It categorizes resources by papers, benchmarks, demos, toolkits, and tutorials, often linking directly to preprints and associated code repositories. The collection targets advancements in agentic reinforcement learning, including methodologies for GUI agents, autonomous research, and strategic tool utilization.