QuantaAlpha is an elite research team dedicated to advancing AI Agent technology, committed to pushing the frontiers of artificial intelligence.
To explore the “quanta” of intelligence and lead the “alpha” frontiers of AI Agents — advancing CodeAgents, self-evolving intelligence, and domain-specialized agents to redefine the boundaries of artificial intelligence research.
Our team currently consists of approximately 20 members, including:
- Professors and associate professors from China's top 985/211 universities
- Postdocs, PhDs, and master's students from prestigious international institutions (CMU, OSU, UTHealth Houston, HKUST, etc.)
- PhD/master's students from China's leading universities (Tsinghua, Peking University, USTC, Zhejiang University, etc.)
Our research focuses on three core areas:
- 💻 CodeAgent: End-to-end execution for real-world coding tasks, enabling agents to autonomously understand, modify, and run complex repositories.
- 📚 DeepResearch: Multi-step reasoning agents for open-ended scientific and knowledge discovery, capable of synthesizing insights across domains.
- 🧠 Agentic Reasoning / Agentic RL: Scaling reinforcement learning with reasoning-centered agents, bridging decision-making, planning, and adaptive intelligence.
- 🔄 Self-Evolving & Co-Learning Agents that continuously improve through self-evolution and collaborative learning, enabling emergent collective intelligence.
Check out our latest research on arXiv:
- GitTaskBench: A Benchmark for Code Agents Solving Real-World Tasks Through Code Repository Leveraging
- CogFSM: Enabling Explicit and Capability-aware Multi-Agent Systems with Self-Learning Finite State Machines
- RepoMaster: Autonomous Exploration and Understanding of GitHub Repositories for Complex Task Solving
- SE-Agent: Self-Evolution Trajectory Optimization in Multi-Step Reasoning with LLM-Based Agents
We are actively seeking collaborators and team members who share our passion for AI Agent technology.
- University/Corporate Lab Resources: API access, GPU resources, and academic influence
- Academic Research Collaboration: Joint research projects and publications
- GitHub Open Source Collaboration: Contributing to and maintaining open-source projects
- Startup Collaboration: Industry partnerships and real-world applications
- Strong Algorithmic Capabilities: Published at least one A-class conference paper as first author (e.g., ACL, NeurIPS, etc.)
- Exceptional Self-Motivation: Ultimate pursuit of technology and academia
- Research Experience: Paper and achievement accumulation in Agent and LLM directions, with GitHub open-source experience
- Research Focus: RL + Agentic Reasoning / CodeAgent / DeepSearch / Those with actual landing projects and ideas are also welcome
- Email: [email protected]
- GitHub: https://github.com/QuantaAlpha
Visit our homepage: https://quantaalpha.github.io
Founded in April 2025