Haodong Lu

University of New South Wales

photo.jpeg

I am Haodong Lu, a PhD candidate at the University of New South Wales (UNSW), Australia, having recently submitted my thesis, under the supervision of Dr. Dong Gong and Prof. Lina Yao, and work closely with CSIRO's Data61 as a scholarship recipient. I am also a recipient of the 2024 Google PhD Fellowship. Before that, I completed both my Honours and BSc. degrees at the University of New South Wales.

My research focuses on Out-of-Distribution Detection and Continual Learning in Computer Vision and Multi-Modal foundation models. I address critical reliability challenges, specifically detecting unseen anomalies and mitigating catastrophic forgetting during sequential post-training. Specializing in prototypical learning, Parameter-Efficient Fine-Tuning (PEFT), and Mixture of Experts (MoE), I have a strong track record of publishing in top-tier venues and am passionate about translating advanced research concepts into scalable, real-world solutions.

Selected Publications

  1. Preprint
    Little By Little: Continual Learning via Self-Activated Sparse Mixture-of-Rank Adaptive Learning
    Haodong Lu, Chongyang Zhao, Jason Xue, Lina Yao, Kristen Moore, Dong Gong
    Preprint. Under Review
    ARXIV
    CVPR
    On Token’s Dilemma: Dynamic MoE with Drift-Aware Token Assignment for Continual Learning of Large Vision Language Models
    Chongyang Zhao, Mingsong Li, Haodong Lu, Dong Gong
    Computer Vision and Pattern Recognition Conference (CVPR), 2026
    ARXIV Project Github
    Preprint
    Adaptive Rank, Reduced Forgetting: Knowledge Retention in Continual Learning Vision-Language Models with Dynamic Rank-Selective LoRA
    Haodong Lu, Chongyang Zhao, Jason Xue, Lina Yao, Kristen Moore, Dong Gong
    Preprint. Under Review
    ARXIV Github
    ICLR
    Learning with Mixture of Prototypes for Out-of-Distribution Detection
    Haodong Lu, Dong Gong, Shuo Wang, Jason Xue, Lina Yao, Kristen Moore
    In the 12th International Conference on Learning Representations (ICLR), 2024.
    ICLR Github
    CVPR
    Self-Expansion of Pre-trained Models with Mixture of Adapters for Continual Learning
    Huiyi Wang, Haodong Lu, Lina Yao, Dong Gong
    Computer Vision and Pattern Recognition Conference (CVPR), 2025
    CVPR Github
    TMLR
    Continual Learning on CLIP via Incremental Prompt Tuning with Intrinsic Textual Anchor
    Haodong Lu, Xinyu Zhang, Kristen Moore, Jason Xue, Lina Yao, Anton van den Hengel, Dong Gong
    Transactions on Machine Learning Research (TMLR), 2025
    TMLR Github
    NeurIPS
    Model Inversion with Layer-Specific Modeling and Alignment for Data-Free Continual Learning
    Rui Tong, Haodong Lu, Yuxuan Liu, Dong Gong
    Advances in Neural Information Processing Systems (NeurIPS), 2025
    NeurIPS Github
    WWW
    Balancing Plasticity and Stability via Dual-Branch Learning in Online Continual Learning
    Xu Han, Haodong Lu, Huiyi Wang, Dong Gong
    The ACM Web Conference (WWW) - short paper track, 2025
    WWW Github

Awards

  • Google PhD Fellowship, Machine Intelligence Track, Google, 2024
  • Outstanding Reviewer, ACM Multimedia, 2024, CVPR, 2025
  • Data61 PhD Scholarship coupled with a Top-up Scholarship, CSIRO’s Data61, 2022
  • Dean’s Honours List, University of New South Wales, 2020

Services

I have served as a reviewer for the following conferences:

  • Conferences: ICML, ICCV, NeurIPS, ICLR, ACMMM, CVPR