Junyuan "Jason" Hong

Junyuan "Jason" Hong

Incoming Assistant Professor

National University of Singapore

I am an incoming Assistant Professor at the ECE department of the National University of Singapore, after spending one year at Massachusetts General Hospital & Harvard Medical School. Previously, I was a postdoctoral fellow advised by Dr. Atlas Wang in the Institute for Foundations of Machine Learning (IFML), affiliated with the UT AI Health Lab and the Good System Challenge, and obtained my Ph.D. in Computer Science and Engineering from Michigan State University with Dr. Jiayu Zhou.

I was recognized as one of the MLSys Rising Stars in 2024 and received a Best Paper Nomination at VLDB 2024. My work was covered by Nature News, The White House, WIRED, Forbes, and FORTUNE.

I lead the Cognitive Science & Trustworthy AI (CoSTA@NUS) Lab exploring the frontier where human minds meet machine intelligence. Openings are available in terms of RA/internship students.

Recent News
UT ECE News Covered by: UT ECE News
MSU Office of Research and Innovation Covered by: MSU Office of Research and Innovation
The White House Covered by: The White House
Awards
Funding
I am grateful that our research is supported by the multiple programs.
Interests
  • Responsible AI
  • Healthcare
  • Privacy
Education

  • PhD in CSE, 2023

    Michigan State University (Advisor: Jiayu Zhou)

    Committee: Anil K. Jain, Sijia Liu, Atlas Wang, Jiayu Zhou

  • MSc in Computer Science, 2018

    University of Science and Technology of China

  • BSc in Physics, minor in CS., 2015

    University of Science and Technology of China

Publications

.js-id-Selected
Selected Healthcare Privacy Trustworthy Large Models FL All
TMLR 2026 POPS: Recovering Unlearned Multi-Modality Knowledge in MLLMs with Prompt-Optimized Parameter Shaking.
Zhangheng Li, Junyuan "Jason" Hong, Jianing Zhu, Sungmin Eum, Shuowen Hu, Suya You, Zhangyang Wang. PDF
FMEA@CVPR 2026 A Physics-Grounded Benchmark for Multi-Agent Dynamics in World Models.
Nuo Chen, Lulin Liu, Zihao Li, Ziyao Zeng, Zihao Zhu, Wenyan Cong, Junyuan "Jason" Hong, Yunhao Yang, Zhengzhong Tu, Yan Wang, Boris Ivanovic, Marco Pavone, Zhangyang Wang, Yang Zhou, Zhiwen Fan. PDF
arXiv 2026 The Last Human-Written Paper: Agent-Native Research Artifacts.
Jiachen Liu, Jiaxin Pei, Jintao Huang, Chenglei Si, Ao Qu, Xiangru Tang, Runyu Lu, Lichang Chen, Xiaoyan Bai, Haizhong Zheng, Carl Chen, Zhiyang Chen, Haojie Ye, Yujuan Fu, Zexue He, Zijian Jin, Zhenyu Zhang, Shangquan Sun, Maestro Harmon, John Dianzhuo Wang, Jianqiao Zeng, Jiachen Sun, Mingyuan Wu, Baoyu Zhou, Yuchen You, Shijian Lu, Yiming Qiu, Fan Lai, Yuan Yuan, Yao Li, Junyuan "Jason" Hong, Ruihao Zhu, Beidi Chen, Alex Pentland, Ang Chen, Mosharaf Chowdhury, Zechen Zhang. PDF Code Blog
CAIS 2026 Scaling Textual Gradients via Sampling-Based Momentum.
Zixin Ding, Junyuan "Jason" Hong, Zhan Shi, Jiachen T. Wang, Zinan Lin, Li Yin, Meng Liu, Zhangyang Wang, Yuxin Chen. PDF
arXiv 2026 CATNIP: LLM Unlearning via Calibrated and Tokenized Negative Preference Alignment.
Zhengbang Yang, Yisheng Zhong, Junyuan "Jason" Hong, Zhuangdi Zhu. PDF
ArXiv 2025 LLMs Can Get "Brain Rot"!.
Shuo Xing, Junyuan "Jason" Hong, Yifan Wang, Runjin Chen, Zhenyu Zhang, Ananth Grama, Zhengzhong Tu, Zhangyang Wang. PDF Code 🌍 Website
ICRA 2026 AD-VF: LLM-Automatic Differentiation Enables Fine-Tuning-Free Robot Planning from Formal Methods Feedback.
Yunhao Yang, Junyuan "Jason" Hong, Gabriel Jacob Perin, Zhiwen Fan, Li Yin, Zhangyang Wang, Ufuk Topcu. PDF
COLM 2025 LoX: Low-Rank Extrapolation Robustifies LLM Safety Against Fine-tuning.
Gabriel J. Perin, Runjin Chen, Xuxi Chen, Nina S. T. Hirata, Zhangyang Wang, Junyuan "Jason" Hong. PDF Code
COLM 2025 More is Less: The Pitfalls of Multi-Model Synthetic Preference Data in DPO Safety Alignment.
Yifan Wang, Runjin Chen, Bolian Li, David Cho, Yihe Deng, Ruqi Zhang, Tianlong Chen, Zhangyang Wang, Ananth Grama, Junyuan "Jason" Hong. PDF
ICML 2025 GuardAgent: Safeguard LLM Agents by a Guard Agent via Knowledge-Enabled Reasoning.
Zhen Xiang, Linzhi Zheng, Yanjie Li, Junyuan "Jason" Hong, Qinbin Li, Han Xie, Jiawei Zhang, Zidi Xiong, Chulin Xie, Carl Yang, Dawn Song, Bo Li. PDF 🏁 Competition
COLM 2025 SEAL: Steerable Reasoning Calibration of Large Language Models for Free.
Runjin Chen, Zhenyu Zhang, Junyuan "Jason" Hong, Souvik Kundu, Zhangyang Wang. PDF Code
EMNLP (Main) 2025 MedHallu: A Comprehensive Benchmark for Detecting Medical Hallucinations in Large Language Models.
Shrey Pandit, Jiawei Xu, Junyuan "Jason" Hong, Zhangyang Wang, Tianlong Chen, Kaidi Xu, Ying Ding. PDF Code πŸ€— Dataset 🌍 Website
FM4Science 2024 DeepOSets: Non-Autoregressive In-Context Learning of Supervised Learning Operators.
Shao-Ting Chiu, Junyuan "Jason" Hong, Ulisses Braga-Neto. PDF
NAACL (Main) 2025 Extracting and Understanding the Superficial Knowledge in Alignment.
Runjin Chen, Gabriel Jacob Perin, Xuxi Chen, Xilun Chen, Yan Han, Nina S. T. Hirata, Junyuan "Jason" Hong, Bhavya Kailkhura. PDF Code
NAACL 2025 GuideLLM: Exploring LLM-Guided Conversation with Applications in Autobiography Interviewing.
Jinhao Duan, Xinyu Zhao, Zhuoxuan Zhang, Eunhye Grace Ko, Lily Boddy, Chenan Wang, Tianhao Li, Alexander Rasgon, Junyuan "Jason" Hong, Min Kyung Lee, Chenxi Yuan, Qi Long, Ying Ding, Tianlong Chen, Kaidi Xu. PDF
VLDB (Best Paper Finalist) 2024 LLM-PBE: Assessing Data Privacy in Large Language Models.
Qinbin Li, Junyuan "Jason" Hong, Chulin Xie, Jeffrey Tan, Rachel Xin, Junyi Hou, Xavier Yin, Zhun Wang, Dan Hendrycks, Zhangyang Wang, Bo Li, Bingsheng He, Dawn Song. PDF Code 🌍 Website 🏁 Competition πŸ† Best Paper Nomination Finetune Code
ICML 2024 Decoding Compressed Trust: Scrutinizing the Trustworthiness of Efficient LLMs Under Compression.
Junyuan "Jason" Hong, Jinhao Duan, Chenhui Zhang, Zhangheng Li, Chulin Xie, Kelsey Lieberman, James Diffenderfer, Brian Bartoldson, Ajay Jaiswal, Kaidi Xu, Bhavya Kailkhura, Dan Hendrycks, Dawn Song, Zhangyang Wang, Bo Li. PDF πŸ€— Models 🌍 Website
ICML 2024 Revisiting Zeroth-Order Optimization for Memory-Efficient LLM Fine-Tuning: A Benchmark.
Yihua Zhang, Pingzhi Li, Junyuan "Jason" Hong, Jiaxiang Li, Yimeng Zhang, Wenqing Zheng, Pin-Yu Chen, Jason D. Lee, Wotao Yin, Mingyi Hong, Zhangyang Wang, Sijia Liu, Tianlong Chen. PDF Code πŸ‘¨β€πŸ«Tutorial
ICLRW 2024 A-CONECT: Designing AI-based Conversational Chatbot for Early Dementia Intervention.
Junyuan "Jason" Hong, Wenqing Zheng, Han Meng, Siqi Liang, Anqing Chen, Hiroko H. Dodge, Jiayu Zhou, Zhangyang Wang. PDF Website πŸ€–Demo
AISTATS 2024 On the Generalization Ability of Unsupervised Pretraining.
Yuyang Deng, Junyuan "Jason" Hong, Jiayu Zhou, Mehrdad Mahdavi. PDF
ICLR 2024 Safe and Robust Watermark Injection with a Single OoD Image.
Shuyang Yu, Junyuan "Jason" Hong, Haobo Zhang, Haotao Wang, Zhangyang Wang, Jiayu Zhou. PDF Code
SaTML 2024 Shake to Leak: Fine-tuning Diffusion Models Can Amplify the Generative Privacy Risk.
Zhangheng Li, Junyuan "Jason" Hong, Bo Li, Zhangyang Wang. PDF Code
ICLR (Spotlight) 2024 DP-OPT: Make Large Language Model Your Privacy-Preserving Prompt Engineer.
Junyuan "Jason" Hong, Jiachen T. Wang, Chenhui Zhang, Zhangheng Li, Bo Li, Zhangyang Wang. PDF Code
NeurIPS-RegML 2023 Who Leaked the Model? Tracking IP Infringers in Accountable Federated Learning.
Shuyang Yu, Junyuan "Jason" Hong, Yi Zeng, Fei Wang, Ruoxi Jia, Jiayu Zhou. PDF
NeurIPS 2023 Understanding Deep Gradient Leakage via Inversion Influence Functions.
Haobo Zhang, Junyuan "Jason" Hong, Yuyang Deng, Mehrdad Mahdavi, Jiayu Zhou. PDF Code
KDDW 2023 A Privacy-Preserving Hybrid Federated Learning Framework for Financial Crime Detection.
Haobo Zhang, Junyuan "Jason" Hong, Fan Dong, Steve Drew, Liangjie Xue, Jiayu Zhou. PDF Code πŸ† PETs πŸ›οΈ White House
KDDW 2023 FedNoisy: A Federated Noisy Label Learning Benchmark.
Siqi Liang, Jintao Huang, Junyuan "Jason" Hong, Fan Dong, Dun Zeng, Jiayu Zhou, Zenglin Xu. PDF Code
ICML 2023 Revisiting Data-Free Knowledge Distillation with Poisoned Teachers.
Junyuan "Jason" Hong, Yi Zeng, Shuyang Yu, Lingjuan Lyu, Ruoxi Jia, Jiayu Zhou. PDF Code Poster
TMLR 2023 How Robust is Your Fairness? Evaluating and Sustaining Fairness under Unseen Distribution Shifts.
Haotao Wang, Junyuan "Jason" Hong, Jiayu Zhou, Zhangyang Wang. PDF
ICLR 2023 MECTA: Memory-Economic Continual Test-Time Model Adaptation.
Junyuan "Jason" Hong, Lingjuan Lyu, Jiayu Zhou, Michael Spranger. PDF Code Slides
ICLR (Spotlight) 2023 Turning the Curse of Heterogeneity in Federated Learning into a Blessing for Out-of-Distribution Detection.
Shuyang Yu, Junyuan "Jason" Hong, Haotao Wang, Zhangyang Wang, Jiayu Zhou. PDF Code
AAAI (Oral) 2023 Federated Robustness Propagation: Sharing Adversarial Robustness in Federated Learning.
Junyuan "Jason" Hong, Haotao Wang, Zhangyang Wang, Jiayu Zhou. Preprint Code Poster
Preprint 2022 Precautionary Unfairness in Self-Supervised Contrastive Pre-training.
Junyuan "Jason" Hong, Haotao Wang, Haobo Zhang, Zhangyang Wang, Jiayu Zhou. Preprint
NeurIPS 2022 Outsourcing Training without Uploading Data via Efficient Collaborative Open-Source Sampling.
Junyuan "Jason" Hong, Lingjuan Lyu, Jiayu Zhou, Michael Spranger. PDF Poster Slides
NeurIPS 2022 Trap and Replace: Defending Backdoor Attacks by Trapping Them into an Easy-to-Replace Subnetwork.
Haotao Wang, Junyuan "Jason" Hong, Aston Zhang, Jiayu Zhou, Zhangyang Wang. PDF Code
ICML 2022 Resilient and Communication Efficient Learning for Heterogeneous Federated Systems.
Zhuangdi Zhu, Junyuan "Jason" Hong, Steve Drew, Jiayu Zhou. PDF
FAccT 2022 Dynamic Privacy Budget Allocation Improves Data Efficiency of Differentially Private Gradient Descent.
Junyuan "Jason" Hong, Zhangyang Wang, Jiayu Zhou. Preprint Slides
ICLR 2022 Efficient Split-Mix Federated Learning for On-Demand and In-Situ Customization.
Junyuan "Jason" Hong, Haotao Wang, Zhangyang Wang, Jiayu Zhou. PDF Code Slides Video
KDD 2021 Federated Adversarial Debiasing for Fair and Transferable Representations.
Junyuan "Jason" Hong, Zhuangdi Zhu, Shuyang Yu, Hiroko Dodge, Zhangyang Wang, Jiayu Zhou. PDF Code Slides
ICML 2021 Data-Free Knowledge Distillation for Heterogeneous Federated Learning.
Zhuangdi Zhu, Junyuan "Jason" Hong, Jiayu Zhou. Preprint Code
AAAI 2021 Learning Model-Based Privacy Protection under Budget Constraints.
Junyuan "Jason" Hong, Haotao Wang, Zhangyang Wang, Jiayu Zhou. PDF Slides Video Supplementary
AD 2020 Detecting MCI using real-time, ecologically valid data capture methodology: How to improve scientific rigor in digital biomarker analyses.
Junyuan "Jason" Hong, Jeffrey Kaye, Hiroko H Dodge, Jiayu Zhou. PDF
TKDD 2019 Variant Grassmann Manifolds: A Representation Augmentation Method for Action Recognition.
Junyuan "Jason" Hong, Yang Li, Huanhuan Chen. PDF DOI
TNNLS 2019 Short Sequence Classification Through Discriminable Linear Dynamical System.
Yang Li, Junyuan "Jason" Hong, Huanhuan Chen. DOI
KDD (Oral) 2018 Disturbance Grassmann Kernels for Subspace-Based Learning.
Junyuan "Jason" Hong, Huanhuan Chen, Feng Lin. PDF Cite Video DOI Supplementary
ECML 2016 Sequential Data Classification in the Space of Liquid State Machines.
Yang Li, Junyuan "Jason" Hong, Huanhuan Chen. PDF Code

Experiences

Experiences

Media Coverage

Invited Talks & Guest Lectures

  • ‘Brain Rot in LLMs: When Benign Data Degrades Intelligence?’ @ LockLLM Workshop at NeurIPS 2025, Nov 2025.
  • ‘GenAI-Based Chatbot for Early Dementia Intervention’ @ Rising Star Symposium Series, IEEE TCCN Special Interest Group for AI and Machine Learning in Security, September, 2024: [link]
  • ‘Building Conversational AI for Affordable and Accessible Early Dementia Intervention’ @ AI Health Course, The School of Information, UT Austin, April, 2024: [paper]
  • ‘Shake to Leak: Amplifying the Generative Privacy Risk through Fine-Tuning’ @ Good Systems Symposium: Shaping the Future of Ethical AI, UT Austin, March, 2024: [paper]
  • ‘Foundation Models Meet Data Privacy: Risks and Countermeasures’ @ Trustworthy Machine Learning Course, Virginia Tech, Nov, 2023
  • ‘Economizing Mild-Cognitive-Impairment Research: Developing a Digital Twin Chatbot from Patient Conversations’ @ BABUΕ KA FORUM, Nov, 2023: [link]
  • ‘Backdoor Meets Data-Free Learning’ @ Hong Kong Baptist University, Sep, 2023: [slides]
  • ‘MECTA: Memory-Economic Continual Test-Time Model Adaptation’ @ Computer Vision Talks, March, 2023: [slides] [video]
  • ‘Split-Mix Federated Learning for Model Customization’ @ TrustML Young Scientist Seminars, July, 2022: [link] [video]
  • ‘Federated Adversarial Debiasing for Fair and Transferable Representations’, @ CSE Graduate Seminar, Michigan State University, October, 2021: [slides]
  • ‘Dynamic Policies on Differential Private Learning’ @ VITA Seminars, UT Austin, Sep, 2020: [slides]

Services

Teaching