Haonan Wang

Ph.D. of National University Singapore.

prof_pic.jpg

Hey, thanks for stopping by! :wave:

I am currently pursuing my Ph.D. in Computer Science at National University Singapore (NUS), advised by Prof.Kenji Kawaguchi since 2023. Before that, I received my B.S. in Computer Science and Statistics from University of Illinois at Urbana-Champaign (UIUC), where I was fortunated to be mentored by Prof. Jiawei Han and Prof. Jingrui He.

Research: My research focuses on data-centric and generative artificial intelligence. I focus on scrutinizing high-quality datasets to improve the capacity and reliability of AI models, including diffusion and multimodal language models. My goal is to develop AI that comprehends and complements human behaviors, ensuring that the technology advances in harmony with human values and augments human capabilities.

Feel free to reach out if you’re interested in my research or would like to discuss potential collaborations! You can contact me via email, Twitter (@Haonan_Wang_), or WeChat (whn769452159).:smile:

news

Nov 22, 2024 When Precision Meets Position: BFloat16 Breaks Down RoPE in Long-Context Training s on Arxiv! RoPE is Broken 🤯 because of … BFloat16 — Then, How Can We Improve Long-Context Training? 🤔️ 👉 We propose AnchorAttention, the improved attention for LLMs long-context training.
May 10, 2024 :tada: Our work The Stronger the Diffusion Model, the Easier the Backdoor: Data Poisoning to Induce Copyright Breaches Without Adjusting Finetuning Pipeline has been accepted for an Oral presentation at ICML 2024! Welcome to our session Oral 6x Robustness and Safety in Vienna.
Feb 1, 2024 :art: Our work Can AI Be as Creative as Humans? is on Arxiv!
Oct 27, 2023 Recent work: FairIF: Boosting Fairness in Deep Learning via Influence Functions with Validation Set Sensitive Attributes is accepted by WSDM 2024.
Aug 20, 2023 Recent work: Single-Pass Contrastive Learning Can Work for Both Homophilic and Heterophilic Graph is accepted by TMLR.
Nov 22, 2022 :star2: A Neural Corpus Indexer for Document Retrieval got Outstanding Paper in NeurIPS 2022!

selected publications

  1. NeurIPS
    The Stronger the Diffusion Model, the Easier the Backdoor: Data Poisoning to Induce Copyright Breaches Without Adjusting Finetuning Pipeline
    Haonan Wang, Qianli Shen, Yao Tong, Yang Zhang, and 1 more author
    In NeurIPS 2023 Workshop on Backdoors in Deep Learning - The Good, the Bad, and the Ugly 2023
  2. TMLR
    Single-Pass Contrastive Learning Can Work for Both Homophilic and Heterophilic Graph
    Haonan Wang, Jieyu Zhang, Qi Zhu, Wei Huang, and 2 more authors
    Transactions on Machine Learning Research 2023
  3. NeurIPS
    Deep Active Learning by Leveraging Training Dynamics
    Haonan Wang, Wei Huang, Ziwei Wu, Hanghang Tong, and 2 more authors
    In Advances in Neural Information Processing Systems 2022