• Rethinking LLMs for Science
    GSAI@RUC, Beijing, China, Dec 2025

  • AI for Science: Roadmap and Perspectives from PhysAI and GenAI
    Shanghai AI Lab, Shanghai, China, Dec 2025

  • Rethinking LLMs for Science
    魔塔社区开发者嘉年华, Hangzhou, China, Nov 2025

  • Interpreting Latent Space for Molecular Design and Optimization
    PolyU, Hong Kong, China, Nov 2025

  • AI for Science: Roadmap and Perspectives from PhysAI and GenAI
    PolyU, Hong Kong, China, Nov 2025

  • Leveraging LLM Reasoning in AI for Scientific Discovery
    SYSU, Shenzhen, China, Nov 2025

  • Geometric Pretraining for Molecules, Proteins and Materials
    SYSU, Shenzhen, China, Nov 2025

  • AI for Science: Roadmap and Perspectives from PhysAI and GenAI
    JHU, Baltimore, MD, March 2025

  • Physics-Inspired Geometric Pretraining for Molecule Representation
    AAAI 2025 Tutorial, Philadelphia, PA, USA, Feb 2025
    [Website]

  • Multi-modal Foundation Model for Scientific Discovery: With Applications in Chemistry, Material, and Biology
    AAAI 2025 Tutorial, Philadelphia, PA, USA, Feb 2025
    [Website]

  • Molecule Representation Learning and Discovery: A Perspective from Topology, Geometry, and Textual Description
    Westlake University, July 2024

  • Molecule Representation Learning and Discovery: A Perspective from Topology, Geometry, and Textual Description
    AI for Chemistry and Materials Science (AI4CM) Symposium, March 2024

  • AI for Molecule Discovery with Multi-Modal Knowledge
    Ph.D. Defense, Mila-UdeM, 20th July, 2023
    [Paper]

  • 将深度学习与分子学习结合：从拓扑、几何和文本角度进行解析
    知乎文章 & blog, June 2023

  • Molecule Representation Learning and Discovery: A Perspective from Topology, Geometry, and Textual Description
    University of California, Berkeley, March 2023

  • ProteinDT: A Text-guided Protein Design Framework
    NRC NLP seminar, March 2023

  • Molecule Representation Learning and Discovery: A Perspective from Topology, Geometry, and Textual Description
    M2D2, Feb 2023

  • GraphCG: Unsupervised Discovery of Steerable Factors in Graphs
    LOG, Meetup @ Mila, Dec 2022

  • Self-Supervised Pretraining with 3D Molecular Geometry
    Huawei, July 2022

  • Pre-training Molecular Graph Representation with 3D Geometry – Rethinking Self-Supervised Learning on Structured Data
    AI Times, March 2022

  • Pre-training Molecular Graph Representation with 3D Geometry – Rethinking Self-Supervised Learning on Structured Data
    Graph Reading Group, Mila, Feb 2022

  • Pre-training Molecular Graph Representation with 3D Geometry – Rethinking Self-Supervised Learning on Structured Data
    NLP Reading Group, University of Oxford, Feb 2022

  • Molecular Representation Learning with Limited Data
    Predoc-3 Presentation, Mila-UdeM, Dec 2021