Awesome-LLM-Robotics is a curated bibliography that collects papers, code, and resources on applying large language and multi-modal models to robotics and reinforcement learning.
The repository addresses the challenge of tracking research at the intersection of foundation models and robotics by organizing papers into thematic categories including reasoning, planning, manipulation, navigation, safety, and simulation frameworks. It serves as a reference point for researchers and practitioners seeking to understand how language models and vision-language models are being adapted for embodied AI tasks, from high-level task planning to low-level robot control.
Developers should use this resource when surveying the landscape of LLM applications in robotics, whether building systems that ground language in robotic actions, implementing planning algorithms with language model reasoning, or evaluating safety considerations in embodied AI. The repository is particularly useful for understanding how different research groups approach problems like instruction following, spatial reasoning, and multi-step task execution. It complements rather than replaces reading individual papers, offering a structured entry point into a rapidly evolving research area.
The project actively accepts community contributions through pull requests, with clear formatting guidelines to maintain consistency. The maintainers organize papers in reverse chronological order across multiple research directions, indicating ongoing curation as new work emerges in the field.
