Purpose
awesome-llm-powered-agent is a resource collection for the field of large language model (LLM)-powered agents. The project description mentions that thanks to the capabilities of LLMs in planning, reasoning, and tool calling, people are actively researching and developing LLM-powered agents, which have the potential to solve complex tasks autonomously or collaboratively, or simulate human interaction. The goal of this project is to build a comprehensive collection of excellent resources related to LLM agents, covering papers, code repositories, and more, and to strive to keep it regularly and continuously updated.
Main Capabilities
This repository is mainly a categorized and organized index of research resources, containing the following content sections:
- Papers are divided into multiple subcategories:
- Autonomous Task Solver
- General Reasoning & Planning & Tool Using
- Multi-Agent Cooperation
- Framework & Open-Source
- Application
- Web Agents
- RL Agents
- Robotics & Embodied AI
- Gaming & Role-Playing
- Other Applications
- Trustworthy
- Human Interaction Simulation
- Human-Agent Interaction
- Agents-Powered LLMs
- Benchmark
- Survey & Tutorial
- Open-Source Projects: including Autonomous Task Solver Projects and Multi-Agent Simulation Projects).
- Other Related Sources, Acknowledgement.
Repository marker legend: 🔥 indicates papers cited more than 50 times or repositories with more than 200 stars; 📖 indicates papers accepted by well-known conferences/journals.
How to Use
Users can browse and search the papers and project links under the above categories by visiting the GitHub repository page. Each entry usually includes the title, authors, publication time, conference/journal (if any), and links such as paper, code, and project page. The repository encourages contributing resources via PR, issue, email, or other means. If you wish to have your own paper included, you can open a pull request.
Applicable Scenarios
- Researchers looking for related papers and code implementations in the field of LLM agents.
- Developers learning about existing open-source agent frameworks and projects.
- Students or enthusiasts systematically studying the research directions of LLM agents.
Limitations
The repository description explicitly states that the repository is not actively maintained, mainly includes papers that appeared before October 2023, and contains a small number of subsequent papers. Therefore, some of the latest developments may not be included. The repository itself does not provide a code runtime environment or tool functionality; it is only a resource list.
