Mengyue Yang

University of Bristol

prof_pic.jpg

I’m currently a Lecturer (= US Assistant Professor) at the School of Engineering Mathematics at the University of Bristol. I studied for my PhD in the Department of Computer Science at University College London (UCL), advised by Professor Jun Wang.

My research interests are Reinforcement Learning, Multi-agent Systems and Causality, especially causality in world models. If you would like to chat about potential collaboration or common interests, feel free to reach me at yangmy6750@gmail.com.

news

Dec 7, 2024 We are organizing the ICLR 2025 Workshop on World Models: Understanding, Modelling and Scaling. 📜 Submission deadline: February 03, 2025 (4-9 pages) 💻 Website.
Oct 31, 2024 We will give a tutorial for Causality and Large Models at ACML 2024 (Dec 2024). Website. During the tutorial, I will discuss the part of Causality: Agents and Foundation Models.
Aug 10, 2024 We are organizing the First Workshop on Causality and Large Models (C♥️LM) at #NeurIPS2024. 📜 Submission deadline: September 06 (4-6 pages) 💻 Website.
Jul 7, 2024 Invited talk about Causality Agents & Foundation Model at 集智 因果读书会. This talk explores the research line of Causal Agents & Foundation Models and future prospects of the Causal Agent problem. Please refer to the slides for details.
Jul 4, 2024 Invited talk at Causality Conference PCIC 2024. PCIC 2024 is scheduled to be held concurrently with the World Artificial Intelligence Conference (WAIC). My talk is about Essential Causal Representation Learning via the Probability of Sufficient and Necessary Causes. Slides

selected publications

2024

  1. preprint
    Attaining Humans Desirable Outcomes in Human-AI Interaction via Structural Causal Games
    Anjie Liu, Jianhong Wang, Haoxuan Li, Xu Chen, Jun Wang, Samuel Kaski, and Mengyue Yang
    arXiv preprint arXiv:2405.16588, 2024

2023

  1. NeurIPS
    Invariant Learning via Probability of Sufficient and Necessary Causes
    Mengyue Yang, Zhen Fang, Yonggang Zhang, Yali Du, Furui Liu, Jean-Francois Ton, Jianhong Wang, and 1 more author
    NeurIPS (Spotlight 3.06%), 2023

2021

  1. CVPR
    Causalvae: Disentangled representation learning via neural structural causal models
    Mengyue Yang, Furui Liu, Zhitang Chen, Xinwei Shen, Jianye Hao, and Jun Wang
    CVPR, 2021
  2. CIKM
    Top-n recommendation with counterfactual user preference simulation
    Mengyue Yang*, Quanyu Dai*, Zhenhua Dong, Xu Chen, Xiuqiang He, and Jun Wang
    CIKM, 2021