Yao Qin

Assistant Professor @ UC Santa Barbara

Co-Director @ REAL AI initiative

Senior Research Scientist @ Google DeepMind

I am an Assistant Professor at the Department of Electrical and Computer Engineering, affiliated with the Department of Computer Science at UC Santa Barbara, where I am also co-leading the REAL AI initiative. Meanwhile, I am a senior Research Scientist at Google DeepMind, working on Gemini Safety. I obtained my PhD degree at UC San Diego in Computer Science, advised by Prof. Garrison W. Cottrell. During my PhD, I was very fortunate to intern under the supervision of Geoffrey Hinton, Ian Goodfellow and many others.

My research interests primarily focus on robustness in machine learning, such as adversarial robustness, out-of-distribution generalization, and fairness. In addition, I am highly passionate about developing reliable AI-driven models tailored for healthcare, with a particular focus on diabetes management. In my lab, we explore various research themes for designing robust & fair machine learning models. Specifically, our research themes include:

  • AI safety in multi-modality models
  • AI for healthcare, particularly for diabetes
  • Services: Invited to serve as an Area Chair for ICML-24, ICLR-24/23, ICCV-23, Co-Local Arrangement Chair for KDD-2023.

    Hiring! I am actively seeking motivated postdoctoral researchers and students who are enthusiastic about robustness and diabetes care to join my research lab. In addition, we also have open positions for visiting students interested in engaging in collaborative research. You are welcome to contact me at yaoqin@ucsb.edu with your resume if you have an interest to join us. If you are a current student at UCSB, please email me with [UCSB Student] in the title!

    NEWS!
    • July 2024: New work on benchmarking LLMs for nutrition estimation based on meal descriptions.
    • July 2024: New work on model degradation in the chain of diffusion models.
    • May 2024: One paper on out-of-distribution detection is accepted to ICML-2024.
    • May 2024: One paper on Ising Machines for generative AI is accepted to Nature Electronics.
    • Apr. 2024: Two abstracts on understanding glycemic effects of exercise for Type 1 diabetes are accepted to ADA-2024.
    • Apr. 2024: Honored to receive the UCSB Regents’ Junior Faculty Fellowship Award.
    • Apr. 2024: We are organizing ECE Summit Day at UCSB.
    • Feb. 2024: One paper on adversarial transfer learning is accepted to CVPR-2024.
    • Jan. 2024: One paper on improving out-of-distribution robustness is accepted to AISTATS-2024.
    • Jan. 2024: One paper on small medical language models is accepted to ICLR-2024.
    • Dec. 2023: We are organizing Robustness of Few-shot and Zero-shot Learning in Foundation Models at NeurIPS-2023.
    • Oct. 2023: Honored to be awarded by Helmsley Charitable Trust working on exercise specific AID algorithm for Type 1 diabetes.
    • Oct. 2023: We are organizing Responsible Machine Learning Summit at UCSB.
    • Oct. 2023: One paper is accepted to Findings of EMNLP-2023.
    • Sep. 2023: One paper is accepted to NeurIPS-2023.
    • Aug. 2023: We are organizing Southern California Data Science Day at KDD-2023.
    • Jun. 2023: Honored to be awarded the UCSB Faculty Research Grant.
    • Feb. 2023: Invited talk at Information Theory and Applications, 2023.

    Lab Members


    Recommendations! If you are an undergraduate/master student at UCSB and wants to work with me, the best way is to reach out to my amazing PhD students first to seek potential collaborations!

    Mehak Dhaliwal (PhD)
    Working on Large Language Model & AI for Diabetes

    Andong Hua (PhD)
    Working on Robustness in Multimodal & Large Language Model

    Kenan Tang (PhD)
    Working on AI for Diabetes

    Youngseok Yoon (PhD, co-advised with Prof. Haewon Jeong)
    Working on Generative Models

    Preprints


    NutriBench: A Dataset for Evaluating Large Language Models in Carbohydrate Estimation from Meal Descriptions
    Andong Hua*, Mehak Preet Dhaliwal*, Ryan Burke, Yao Qin
    [Paper][Project Page][Data]

    ReDiFine: Reusable Diffusion Finetuning for Mitigating Degradation in the Chain of Diffusion
    Youngseok Yoon, Dainong Hu, Iain Weissburg, Yao Qin, Haewon Jeong
    [Paper]

    Detecting Out-of-Distribution through the Lens of Neural Collapse
    Litian Liu and Yao Qin
    [Paper]

    Towards Robust Prompts on Vision-Language Models
    Jindong Gu, Ahmad Beirami, Xuezhi Wang, Alex Beutel, Philip Torr and Yao Qin
    [Paper]

    Deflecting Adversarial Attacks
    Yao Qin, Nicholas Frosst, Colin Raffel, Garrison Cottrell and Geoffrey Hinton
    [Paper]

    Evaluation Methodology for Attacks Against Confidence Thresholding Models
    Ian Goodfellow, Yao Qin, David Berthelot
    [Paper]

    Selected Publications

    (* indicates equal contributions.)

    Fast Decision Boundary based Out-of-Distribution Detector
    Litian Liu and Yao Qin
    International Conference on Machine Learning (ICML), 2024.
    [Paper][Code]

    Initialization Matters for Adversarial Transfer Learning
    Andong Hua, Jindong Gu, Zhiyu Xue, Nicholas Carlini, Eric Wong and Yao Qin
    IEEE Conference on Computer Vision and Pattern Recognition (CVPR), 2024.
    [Paper][Code]

    Effective Robustness against Natural Distribution Shifts for Models with Different Training Data
    Zhouxing Shi, Nicholas Carlini, Ananth Balashankar, Ludwig Schmidt, Cho-Jui Hsieh, Alex Beutel and Yao Qin
    Advances in Neural Information Processing Systems (NeurIPS), 2023
    [Paper]

    Understanding and Improving Robustness of Vision Transformers through Patch-based Negative Augmentation
    Yao Qin, Chiyuan Zhang, Ting Chen, Balaji Lakshminarayanan, Alex Beutel and Xuezhi Wang
    Advances in Neural Information Processing Systems (NeurIPS), 2022
    [Paper]

    Are Vision Transformers Robust to Patch Perturbations?
    Jindong Gu, Volker Tresp, Yao Qin
    European Conference on Computer Vision (ECCV), 2022
    [Paper][Code]

    Improving Calibration through the Relationship with Adversarial Robustness
    Yao Qin, Xuezhi Wang, Alex Beutel, Ed H. Chi
    Advances in Neural Information Processing Systems (NeurIPS), 2021
    [Paper]

    Detecting and Diagnosing Adversarial Images with Class-Conditional Capsule Reconstructions
    Yao Qin*, Nicholas Frosst*, Sara Sabour, Colin Raffel, Garrison Cottrell and Geoffrey Hinton
    International Conference on Learning Representations (ICLR), 2020
    [Paper]

    Imperceptible, Robust and Targeted Adversarial Examples for Automatic Speech Recognition
    Yao Qin, Nicholas Carlini, Ian Goodfellow, Garrison Cottrell, Colin Raffel
    International Conference on Machine Learning (ICML), 2019.
    [Paper][Project Page][Code]

    Autofocus Layer for Semantic Segmentation
    Yao Qin, Konstantinos Kamnitsas, Siddharth Ancha, Jay Nanavati, Garrison Cottrell, Antonio Criminisi, Aditya Nori
    International Conference on Medical Image Computing and Computer Assisted Intervention (MICCAI), 2018.
    Oral presentation (4% acceptance rate)
    [Paper][Code]

    Hierarchical Cellular Automata for Visual Saliency
    Yao Qin*, Mengyang Feng*, Huchuan Lu, Garrison Cottrell
    International Journal of Computer Vision (IJCV), 2017.
    [Paper][Code]

    A Dual-Stage Attention-Based Recurrent Neural Network for Time Series Prediction
    Yao Qin, Dongjin Song, Haifeng Chen, Wei Cheng, Guofei Jiang, Garrison Cottrell
    International Joint Conference on Artificial Intelligence (IJCAI), 2017.
    [Paper][Code][Data]

    Saliency Detection via Cellular Automata
    Yao Qin, Huchuan Lu , Yiqun Xu, He Wang
    IEEE Conference on Computer Vision and Pattern Recognition (CVPR), 2015.
    [Paper][Code][中文版]

    Trip Journals

    Movie Reviews


    Other than research, I am also a big fan of old movies. Unfortunately, all my reviews are only in Chinese, but you can use LLM to translate them if you want :)