Add six papers on LLM-based agents and multi-agent systems

[Liu-Hy]

Thank you for maintaining this excellent resource for the community. In this PR, we request to add six recent papers from our lab that are highly relevant to LLM-based agents.

Papers Added

Section 2.1.2 (Innovation-oriented Deployment)

1. GenoTEX (2024/06): An LLM agent benchmark for automated gene expression data analysis. Accepted as an Oral paper to MLCB 2025.

   • Paper | Code | Website

2. GenoMAS (2025/07): A multi-agent framework with 6 specialized LLM agents for scientific discovery via code-driven gene expression analysis.

   • Paper | Code | Website

3. aiXiv (2025/08): A next-generation open-access platform for AI-generated scientific research with multi-agent architecture.

   • Paper | Code | Website

Section 2.2.1 (Ordered Cooperation)

4. CoMAS (2025/10): Co-evolving multi-agent systems via interaction rewards for autonomous agent improvement.
   • Paper

Section 4.1 (Benchmarks for LLM-based Agents)

5. Achilles Heel (2025/04): Identifying trustworthiness challenges in distributed multi-agent systems.
   • Paper

Section 4.2 (Training and Optimizing LLM-based Agents)

6. The Landscape of Agentic RL (2025/09): A comprehensive survey on agentic reinforcement learning for LLMs.
   • Paper | Code

Contribution

These papers contribute to advancing the LLM agent field in several key areas:

• Scientific Discovery: GenoTEX, GenoMAS, and aiXiv demonstrate LLM agents' capabilities in AI4Science
• Multi-Agent Collaboration: GenoMAS and CoMAS showcase structured multi-agent frameworks
• Agent Training: The Agentic RL survey and CoMAS advance training methodologies
• Trustworthiness: Achilles Heel addresses critical security and reliability concerns

All papers follow the repository's format and are placed in appropriate sections based on their primary contributions.

Thank you for considering our request!