Skip to content
@TsinghuaC3I

TsinghuaC3I

Center for Collaborative & Conversational Intelligence at Tsinghua University

The Center for Collaborative & Conversational Intelligence (C3I) is affiliated with the Department of Electronic Engineering at Tsinghua University and is led by Professor Bowen Zhou.

Our research objectives are as follows:

  1. Specialization and Alignment of Foundation Models: This involves in-depth research on the architecture design, training, and alignment of foundation models, exploring specialization in reasoning.
  2. Multi-Dimensional Collaborative Research: Based on a cognitive collaboration framework incorporating both System 1 and System 2, this area explores collaboration between large and small models, multimodal collaboration, and human-machine collaboration. It also investigates the interpretable mechanisms of neurodynamics and theories of sustainable hybrid human-machine intelligence.
  3. Application Development for Scientific Discovery: Focused on the biomedical field, this research utilizes collaborative interaction technologies to drive knowledge innovation and promote scientific discoveries and technological advancements.

Popular repositories Loading

  1. Awesome-RL-for-LRMs Awesome-RL-for-LRMs Public

    A Survey of Reinforcement Learning for Large Reasoning Models

    1k 62

  2. MARTI MARTI Public

    A Framework for LLM-based Multi-Agent Reinforced Training and Inference

    Python 245 26

  3. SSRL SSRL Public

    SSRL: Self-Search Reinforcement Learning

    Python 132 6

  4. MedXpertQA MedXpertQA Public

    [ICML 2025] MedXpertQA: Benchmarking Expert-Level Medical Reasoning and Understanding

    Python 116 4

  5. Unify-Post-Training Unify-Post-Training Public

    Towards a Unified View of Large Language Model Post-Training

    Python 111 4

  6. UltraMedical UltraMedical Public

    [NeurIPS 2024 D&B Track, Spotlight] UltraMedical: Building Specialized Generalists in Biomedicine

    Python 93 4

Repositories

Showing 10 of 18 repositories

Top languages

Python

Most used topics

Loading…