Yawei Li

Affiliation. Email.

yawei_photo.JPG

Institut für Statistik

LMU München

Ludwigstraße 33, 80539

München, Germany

My academic journey started in China, where I obtained a bachelor in Laboratory Medical Science from Southern Medical University in Guangzhou, China. However, my passion for engineering led me to pursue a B.Sc. in Electrical Engineering at TU Darmstadt in Germany. Intrigued by the intersection of technology and medicine, I then completed a M.Sc. in Biomedical Computing at Technical University of Munich (TUM).

Currently, I am a Ph.D. student at the Statistical Learning and Data Science (SLDS) group at LMU Munich, under the supervision of Prof. Dr. Bernd Bischl. I am also a junior member of Munich Center for Machine Learning (MCML). In 2024, I was also a visiting researcher at the Deep Learning Lab of Prof. Dr. Kenji Kawaguchi at the National University of Singapore (NUS).

My research interests lie in trustworthy deep learning, with a particular focus on explainability and uncertainty quantification.

news

Nov 01, 2024 Our paper A Dual-Perspective Approach to Evaluating Feature Attribution Methods is accepted by Transactions on Machine Learning Research (TMLR). (Camera-ready version coming soon)
Oct 09, 2024 Our paper FinerCut: Finer-grained Interpretable Layer Pruning for Large Language Models is accepted by Compression Workshop at NeurIPS 2024.
Feb 09, 2024 The paper Probabilistic Self-supervised Learning via Scoring Rules Minimization is accepted by ICLR 2024.
Oct 28, 2023 The paper AttributionLab: Faithfulness of Feature Attribution Under Controllable Environments was accepted by NeurIPS 2023 XAIA Workshop. For those interested in delving into the details of our work, we have uploaded an extended version (with 9-page main text) on ArXiv.
Jun 22, 2022 The paper Analyzing the Effects of Handling Data Imbalance on Learned Features from Medical Images by Looking Into the Models was accepted by the ICML 2022 Workshop: Interpretable Machine Learning in Healthcare. Check out the ArXiv version.

selected publications

  1. A Dual-Perspective Approach to Evaluating Feature Attribution Methods
    Yawei Li*, Yang Zhang*, Kenji Kawaguchi, and 3 more authors
    Transactions on Machine Learning Research (TMLR), 2024
  2. FinerCut: Finer-grained Interpretable Layer Pruning for Large Language Models
    Yang Zhang*Yawei Li*, Xinpeng Wang, and 5 more authors
    Compression Workshop at NeurIPS 2024, 2024
  3. Probabilistic Self-supervised Representation Learning via Scoring Rules Minimization
    Amirhossein Vahidi, Simon Schoßer, Lisa Wimmer, and 4 more authors
    In The Twelfth International Conference on Learning Representations (ICLR), 2024
  4. AttributionLab: Faithfulness of Feature Attribution Under Controllable Environments
    Yang Zhang*Yawei Li*, Hannah Brown, and 5 more authors
    NeurIPS 2023 Workshop XAI in Action, 2023
  5. Deep learning-based classification of dermatological lesions given a limited amount of labelled data
    S Krammer, Y Li, N Jakob, and 6 more authors
    Journal of the European Academy of Dermatology and Venereology, 2022
  6. Fine-Grained Neural Network Explanation by Identifying Input Features with Predictive Information
    Yang* Zhang, Ashkan* Khakzar, Yawei Li, and 3 more authors
    Advances in Neural Information Processing Systems (NeurIPS), 2021
  7. Explaining covid-19 and thoracic pathology model predictions by identifying informative input features
    Ashkan* Khakzar, Yang* Zhang, Wejdene Mansour, and 5 more authors
    In International Conference on Medical Image Computing and Computer-Assisted Intervention (MICCAI), 2021