Kuangshi (Leo) Ai

University of Notre Dame

prof_pic.jpg

247 Fitzpatrick Hall

Notre Dame, IN 46556

Hi! I’m a third-year CS Ph.D. student at the ND-VIS lab, advised by Prof. Chaoli Wang. I received my bachelor’s degree in AI from Fudan University. Currently, I am a ML Research Scientist Intern at TikTok in San Jose, where I work on long-horizon self-evolving agents for AIGC video editing.

My research lies at the intersection of LLM agents, scientific visualization, and human-computer interaction. I build agentic systems that make complex 3D/4D scientific data easier to explore, understand, and manipulate. My current work focuses on self-evolving agents for long-horizon tasks, particularly through memory, reflection, adaptive learning, and agentic post-training. I also develop benchmarks and evaluation methods for agents. My work has been recognized with an IEEE VIS Best Paper Award.

More broadly, I am interested in Agentic AI + VIS + HCI, including self-evolving agents, 3D interaction, mixed-initiative systems, and human–agent collaboration.

If you’re interested in my work or would like to collaborate, feel free to reach out at kai[at]nd[dot]edu.

news

Jul 16, 2026 Two first-author full papers and one short paper accepted to IEEE VIS 2026!
Jan 28, 2026 🚀 Calling for Collaborators: SciVisAgentBench!
🔗 Learn more or contribute through our project page.
Aug 03, 2025 🏆 IEEE VIS 2025 Best Paper Award!
My first first-author paper NLI4VolVis has been selected as one of only 5 Best Papers out of 537 submissions! 📰 The work was also featured by Notre Dame News.
Jul 15, 2025 Two papers accepted to IEEE VIS 2025! This marks my first first-author paper during my Ph.D.! 🎉
May 16, 2025 One paper accepted to ACL 2025 findings!

selected publications

  1. IEEE VIS
    nli4volvis.png
    🏆 Best Paper VIS 2025 NLI4VolVis: Natural Language Interaction for Volume Visualization via LLM Multi-Agents and Editable 3D Gaussian Splatting
    Kuangshi Ai, Kaiyuan Tang, and Chaoli Wang
    In Proceedings of IEEE Transactions on Visualization and Computer Graphics (IEEE VIS), 2025
    arXiv Video Blog Code Slides Website
  2. IEEE VIS
    scivisagentbench.png
    SciVisAgentBench: A Benchmark for Evaluating Scientific Data Analysis and Visualization Agents
    Kuangshi Ai, Haichao Miao, Kaiyuan Tang, Nathaniel Gorski, Jianxin Sun, Guoxi Liu, Helgi I. Ingolfsson, David Lenz, Hanqi Guo, Hongfeng Yu, Teja Leburu, Michael Molash, Bei Wang, Tom Peterka, Chaoli Wang, and Shusen Liu
    In Proceedings of IEEE Transactions on Visualization and Computer Graphics (IEEE VIS), 2026
    arXiv Code Website HF
  3. IEEE VIS
    hilsva.png
    HiLSVA: Design and Evaluation of a Human-in-the-Loop Agentic System for Scientific Visualization
    Kuangshi Ai, Patrick Phuoc Do, and Chaoli Wang
    In Proceedings of IEEE Transactions on Visualization and Computer Graphics (IEEE VIS), 2026
    arXiv Video Code Website
  4. IEEE VIS
    texgs.png
    TexGS-VolVis: Expressive Scene Editing for Volume Visualization via Textured Gaussian Splatting
    Kaiyuan Tang, Kuangshi Ai, Jun Han, and Chaoli Wang
    In Proceedings of IEEE Transactions on Visualization and Computer Graphics (IEEE VIS), 2025
    arXiv
  5. IEEE VIS Short
    interaction_paradigm.png
    Exploring LLM Agent Designs and Interaction Modalities for Scientific Visualization
    Jackson Vonderhorst, Kuangshi Ai, Haichao Miao, Shusen Liu, and Chaoli Wang
    In Proceedings of IEEE Visualization and Visual Analytics (IEEE VIS Short Paper), 2026
    arXiv
  6. Preprint
    vis_coscientist.png
    Toward AI VIS Co-Scientists: A General and End-to-End Agent Harness for Solving Complex Data Visualization Tasks
    Haichao Miao, Zhimin Li, Kuangshi Ai, Kaiyuan Tang, Chaoli Wang, Peer-Timo Bremer, and Shusen Liu
    2026
    arXiv Video
  7. VIS x GenAI Oral
    eval_agent.png
    An Evaluation-Centric Paradigm for Scientific Visualization Agents
    Kuangshi Ai, Haichao Miao, Zhimin Li, Chaoli Wang, and Shusen Liu
    In Proceedings of the 1st Workshop on GenAI, Agents, and the Future of VIS (IEEE VIS Workshop), 2025
    arXiv Slides
  8. ACL Findings
    variation-theory.png
    Leveraging Variation Theory in Counterfactual Data Augmentation for Optimized Active Learning
    Simret Araya Gebreegziabher, Kuangshi Ai, Zheng Zhang, Elena L. Glassman, and Toby Jia-Jun Li
    In Proceedings of Findings of the Association for Computational Linguistics: ACL, 2025
    arXiv Code
  9. Preprint
    svlat.png
    SVLAT: Scientific Visualization Literacy Assessment Test
    Patrick Phuoc Do, Kaiyuan Tang, Kuangshi Ai, and Chaoli Wang
    2026
    arXiv Supp
  10. Preprint
    scivisagentskills.png
    SciVisAgentSkills: Design and Evaluation of Agent Skills for Scientific Data Analysis and Visualization
    Kuangshi Ai, Haichao Miao, Kaiyuan Tang, Shusen Liu, and Chaoli Wang
    2026
    arXiv Code