Research

Three threads, one goal: AI that people can rely on when the stakes are real.

01

Robust and Reliable AI

The real world is noisy: data can be imperfect, labels inaccurate, and prompts ambiguous. We develop methods with theoretical justification for robust pre-training and inference-time steering.

  • Reducing hallucinations in vision-language models via latent space steering

    Sheng Liu, Haotian Ye, Lei Xing, James Zou

    ICLR 2025 · Spotlight

  • In-context vector: making in-context learning more effective and controllable

    Sheng Liu, Haotian Ye, Lei Xing, James Zou

    ICML 2024

  • Adaptive early-learning correction for segmentation from noisy annotations

    Sheng Liu*, Kangning Liu*, Weicheng Zhu, Yiqiu Shen, Carlos Fernandez-Granda

    CVPR 2022 · Oral

  • Early-learning regularization prevents memorization of noisy labels

    Sheng Liu, Jonathan Niles-Weed, Narges Razavian, Carlos Fernandez-Granda

    NeurIPS 2020

02

AI Systems and AI Agents

We build AI software platforms that assist human experts in clinical practice and promote human–AI collaboration, and we optimize AI systems built with large language models.

  • Optimizing generative AI by backpropagating language model feedback

    Mert Yuksekgonul*, Federico Bianchi*, Joseph Boen*, Sheng Liu*, Pan Lu*, Zhi Huang*, Carlos Guestrin, James Zou

    Nature 2025

  • OctoTools: an agentic framework with extensible tools for complex reasoning

    Pan Lu*, Bowen Chen*, Sheng Liu*, Rahul Thapa, Joseph Boen, James Zou

    NAACL KnowledgeNLP 2025 · Best paper award

03

AI for Human Disease and Health

Medicine is high-stakes: accuracy and reliability are paramount. We build models supporting clinical decisions in radiation oncology and Alzheimer's disease, integrating domain knowledge with modern machine learning.

  • Cerebra: A Multidisciplinary Agentic AI Board for Multimodal Dementia Characterization and Risk Assessment

    Sheng Liu*, Long Chen*, Zeyun Zhao, Qinglin Gou, Qingyue Wei, Arjun Masurkar, Kevin M Spiegler, Philip Kuball, Stefania C Bray, Megan Bernath, Deanna R Willis, Jiang Bian, Lei Xing, Eric Topol, Kyunghyun Cho, Yu Huang, Ruogu Fang, Narges Razavian, James Zou

    Under Review

  • Automated radiotherapy treatment planning guided by GPT-4Vision

    Sheng Liu*, Oscar Pastor-Serrano*, et al., James Zou, Lei Xing

    AAPM 2024 · Best in Medical Physics

  • Generalizable deep learning model for early Alzheimer's disease detection from structural MRIs

    Sheng Liu, Arjun V. Masurkar, Henry Rusinek, Jingyun Chen, Ben Zhang, Weicheng Zhu, Carlos Fernandez-Granda, Narges Razavian

    Nature Scientific Reports 2023