Yuanjie Shi

PhD Candidate in Computer Science at Washington State University

Pullman, Washington

yuanjie.shi@wsu.edu

GitHub

I am a PhD candidate in Computer Science at the School of Electrical Engineering and Computer Science, Washington State University, advised by Prof. Yan Yan.

I am on the job market for faculty and research positions in AI and machine learning.

My research develops foundations and systems for reliable AI. I study how to measure reliability, improve it through learning, and guide AI behavior during reasoning and execution. My work connects statistical guarantees with practical methods for building reliable and controllable AI.

Research directions

  • Foundations of reliable AI: I develop statistical guarantees and learning methods for uncertainty quantification, robust prediction, and preference alignment. My work includes informative conformal prediction sets, risk control under distribution shift, and training objectives that improve robustness to label noise and imperfect preference data.

  • Systems for reliable AI: I develop methods for reliable federated learning, LLM reasoning, and agent execution. My work addresses calibration across heterogeneous clients, factuality control, error propagation in LLM pipelines, evidence requirements for tool actions, and risk-triggered intervention in embodied agents.

My work has appeared at ICML, ICLR, NeurIPS, AISTATS, AAAI, and UAI. My honors include the WSU EECS Outstanding Dissertation Award (2026), the WSU EECS Outstanding Research Assistant Award (2025), and the Mahmoud M. Dillsi Family Graduate Fellowship (2023–2024). I have also been recognized as an ICML Gold Reviewer (2026), a UAI Top Reviewer (2026), and a NeurIPS Top Reviewer (2025).

selected publications

  1. LLM, Reliable AI
    neurips2026_reliability_fig1a.png
    A Latent-Load Framework for Reliability Analysis and Intervention Design in LLM Pipelines
    Yuanjie Shi and Yan Yan
    In NeurIPS, 2026
  2. FL, CP
    uai26_sketch_size_effect.png
    Valid and Efficient Uncertainty Quantification for Federated Joint Shift
    Yuanjie Shi, Peihong Li, Xuanyu Cao, and Yan Yan
    In UAI, 2026
  3. LLM, CP
    ICML26_inference_time.png
    Inference-Time Conformal Reasoning with Valid Factuality Control for Large Language Models
    Ting Wang, Yuanjie Shi, Yan Yan, and Huan Zhang
    In ICML, 2026
    Ting Wang and Yuanjie Shi contributed equally.
  4. LLM, Alignment
    ICLR2026.png
    Keep the Best, Forget the Rest: Reliable Alignment with Order-Aware Preference Optimization
    Jiahui Zhu, Yuanjie Shi, Xiyue Peng, Xin Liu, Yan Yan, and Honghao Wei
    In ICLR, 2026
    Jiahui Zhu and Yuanjie Shi contributed equally.
  5. Robust ML, CP
    AISTATS2026.png
    Conformal Margin Risk Minimization: An Envelope Framework for Robust Learning under Label Noise
    Yuanjie Shi, Peihong Li, Zijian Zhang, Jana Doppa, and Yan Yan
    In AISTATS, 2026
    Yuanjie Shi and Peihong Li contributed equally.
  6. ML, CP
    ICML25_DPSM.png
    Direct Prediction Set Minimization via Bilevel Conformal Classifier Training
    Yuanjie Shi, Hooman Shahrokhi, Xuesong Jia, Xiongzhi Chen, Jana Doppa, and Yan Yan
    In ICML, 2025
    Yuanjie Shi and Hooman Shahrokhi contributed equally.
  7. CP
    Conformal Prediction for Class-wise Coverage via Augmented Label Rank Calibration
    Yuanjie Shi, Subhankar Ghosh, Taha Belkhouja, Jana Doppa, and Yan Yan
    In NeurIPS, 2024