This repository serves as a curated paper list for "The Rise and Potential of Large Language Model Based Agents: A Survey," targeting researchers and practitioners in AI and LLM development. It provides a comprehensive overview of LLM-based agents, categorizing research by agent construction, applications, and societal interactions, offering a valuable resource for understanding the state-of-the-art and future directions in this rapidly evolving field.

How It Works

The repository organizes papers based on the core components of LLM-based agents: Brain (LLM capabilities like knowledge, memory, reasoning), Perception (multimodal inputs), and Action (tool use, embodied actions). It further details applications in single-agent, multi-agent, and human-agent scenarios, and explores agent societies, behavior, and environments. This structured approach allows users to navigate the vast literature efficiently.

Quick Start & Requirements

This repository is a curated list of research papers and does not require installation or execution. Links to papers and code are provided within the README for further exploration.

Highlighted Details

• Comprehensive categorization of LLM-based agent research, covering foundational concepts to advanced applications.
• Includes links to papers and code for over 100 research projects, facilitating direct access to implementations and findings.
• Regularly updated with new research, including recent work on AgentGym, R3 (Reverse Curriculum Reinforcement Learning), and multimodal agents.
• Provides a detailed table of contents mirroring the structure of the main survey paper.

Maintenance & Community

The project is maintained by Zhiheng Xi and a team of contributors, with regular updates and additions of new research. Contact information for the primary author is provided.

Licensing & Compatibility

The repository itself is not licensed for software use; it is a collection of links to research papers. The licensing of individual papers and their associated code would need to be checked on their respective sources.

Limitations & Caveats

This repository is a static list of papers and does not provide executable code or a unified framework for building LLM agents. The "coming soon" note about adding one-sentence intros to each paper indicates ongoing development.