Yaxin Du

InfoMosaic-Bench: Evaluating Multi-Source Information Seeking in Tool-Augmented Agents

SWE-Dev: Training and Evaluating Autonomous End-to-End Feature-Driven Software Development

Selecting Auxiliary Data via Neural Tangent Kernels for Low-Resource Domains

MCP-Flow: Facilitating LLM Agents to Master Real-World, Diverse and Scaling MCP Tools

Towards Scalable Web Browsing via Tool-Augmented Programmatic Agent Pair

Self-Evolving Multi-Agent Collaboration Networks for Software Development

MAS-GPT: Training LLMs to Build LLM-based Multi-Agent Systems

Federated Instruction Tuning of LLMs with Domain Coverage Augmentation