Yifei Zhang

Professor, School of Computer Science, Northwestern Polytechnical University

group/yifei-zhang.jpg

No. 127 Youyi Xilu

Beilin District, Xi'an, Shaanxi, China

yifei.zhang@nwpu.edu.cn

I am a Professor in the School of Computer Science at Northwestern Polytechnical University. My research focuses on trustworthy machine learning, federated learning, graph representation learning, large language models, and robust vision-language models.

Before joining NWPU, I was a Research Scientist at Nanyang Technological University, received my Ph.D. from The Chinese University of Hong Kong, and worked as a data scientist and software engineer in industry. My work has appeared in venues including NeurIPS, ICML, ICLR, KDD, CVPR, ICCV, WWW, ACL, EMNLP, AAAI, SIGIR, and CIKM.

I am broadly interested in building reliable learning systems for foundation models with focusing on data-centric machine intelligence, including self-supervised learning, agentic learning, spatial intelligence, representation learning, and knowledge discovery from large-scale multimodal data.

I lead the DISC Lab, the Data Intelligence and Scientific Discovery Lab (DISC Lab) advances the frontier of data intelligence and scientific discovery, transforming complex data into trustworthy insights, intelligent systems, and scientific breakthroughs.

Group News

  • 2026 TRACE appears at WWW 2026. Trajectory-aware evaluation for deep research agents.
  • 2026 Hierarchically Robust Zero-shot Vision-Language Models appears at CVPR 2026. Robust zero-shot vision-language modeling.
  • 2026 Geometric Collapse appears at ICML 2026. Studying when vision models fail to verify physical causality.
  • 2025 CrossSpectra appears at NeurIPS 2025. Cross-layer smoothness for parameter-efficient fine-tuning.

Snapshots from group seminars, student discussions, and academic events.

Group weekly seminar placeholder
Weekly Seminar Research talks, reading groups, and project updates.
Group research discussion placeholder
Research Discussion Small-group brainstorming around trustworthy machine learning problems.
Group academic event placeholder
Academic Events Conferences, workshops, and invited academic exchanges.
Group student life placeholder
Student Life Shared moments from group gatherings and student activities.

selected publications

  1. ICML
    Geometric Collapse: When Vision Models Fail to Verify Physical Causality
    Wentao Zhang, Jinhu Qi, Weiqiang Jin, and 3 more authors
    In International Conference on Machine Learning, 2026
  2. CVPR
    Hierarchically Robust Zero-shot Vision-Language Models
    Junhao Dong, Yifei Zhang, Hao Zhu, and 2 more authors
    In Conference on Computer Vision and Pattern Recognition, 2026
  3. WWW
    TRACE: Trajectory-Aware Comprehensive Evaluation for Deep Research Agents
    Yanyu Chen, Jiyue Jiang, Jiahong Liu, and 3 more authors
    In The Web Conference, 2026
  4. NeurIPS
    CrossSpectra: Exploiting Cross-Layer Smoothness for Parameter-Efficient Fine-Tuning
    Yifei Zhang, Hao Zhu, Junhao Dong, and 4 more authors
    In Conference on Neural Information Processing Systems, 2025
  5. KDD
    Understanding and Mitigating Hyperbolic Dimensional Collapse in Graph Contrastive Learning
    Yifei Zhang, Hao Zhu, Mengling Yang, and 4 more authors
    In SIGKDD Conference on Knowledge Discovery and Data Mining, 2025
  6. CVPR
    pFedMixF: Personalized Federated Class-Incremental Learning with Mixture of Frequency Aggregation
    Yifei Zhang, Hao Zhu, Alysa Ziying Tan, and 3 more authors
    In Conference on Computer Vision and Pattern Recognition, 2025
  7. KDD
    Geometric View of Soft Decorrelation in Self-Supervised Learning
    Yifei Zhang, Hao Zhu, Zixing Song, and 4 more authors
    In SIGKDD Conference on Knowledge Discovery and Data Mining, 2024
  8. NeurIPS
    Mitigating the Popularity Bias in Graph Collaborative Filtering: A Dimensional Collapse Perspective
    Yifei Zhang, Hao Zhu, Yankai Chen, and 3 more authors
    In Conference on Neural Information Processing Systems, 2023