About Me
I am currently a Ph.D. student in Computer Science at North Carolina State University, where I am fortunate to be advised by Prof. Xiaorui Liu. Prior to that, I received my M.S. degree in Mathematics from the Chinese Academy of Sciences, and my B.S. degree in Mathematics from Beijing Normal University. I have gained diverse industry research experience at Trustworthy Shopping at Amazon (2026), Automated Reasoning at AWS AI (2025), Amazon GuardDuty (2024), Baidu, Inc. (2023), and Tsinghua AIR (2022).
My research focuses on building trustworthy and efficient AI systems that extend naturally to real-world applications in science and industry:
▶ Trustworthy AI
- Design universal robustness-informed architectures, including robust transformers [ProTransformer, NeurIPS'24], reprogrammable representations that adapt any pretrained deep learning model for robustness [NRPM, ICLR'25], robust aggregation for graph learning [RUNG, NeurIPS'24] [BBRW, TKDD'26], robust dictionary learning [EDLNets], and noise-conditional networks [TCN].
- Achieve better accuracy-robustness Pareto frontiers through robustness reprogramming under three paradigms [Robustness Reprogramming, ICLR'25], noise-conditional networks [TCN], and mixtures of robust experts.
- Develop efficient iterative attacks and adversarial training [Spiking Attacks, ICLR'26].
▶ Efficient AI
- Address efficiency problems in safety and robustness: efficient iterative attacks and adversarial training [Spiking Attacks, ICLR'26], training-free robust deployment [ProTransformer, NeurIPS'24] [Robustness Reprogramming, ICLR'25], one-for-all defense [TCN], and efficient certifiable defense [SRS, NeurIPS'24].
- Improve efficiency by reducing redundancy in sequential modeling: accelerating iterative attacks [Spiking Attacks, ICLR'26], accelerating randomized smoothing [SRS, NeurIPS'24], and accelerating generative modeling via modulated quantization [ModDiff, ICML'25].
- Pursue context- and token-efficiency in LLMs: a hierarchical transformer for log data [HLogformer], and context-efficient modeling for structured data.
▶ AI for Science & Industry
- Biology: molecular dynamics modeling [ESTAG, NeurIPS'23] and molecule generation [MolConf, ICLR'26]; genomics, including explainable survival analysis [PathExpSurv, BMC'23], LLMs for lncRNA regulation [LLM4Bio, MABM'25], single-cell chromatin accessibility [SCARP, BiB'24], and interpretable scientific interaction discovery [XAI4Sci, ICLR WS'25].
- Interdisciplinary AI: extending these methods to networking and transportation.
- LLM post-training for industrial applications: natural language formalization [SCD] and context-efficient modeling [HLogformer].
News
June, 2026
Our paper on Robustness in Directed Graphs is accepted by ACM Transactions on Knowledge Discovery from Data (TKDD).
May, 2026
The Neurosymbolic LLM Reasoning proposal I led was honored with an Amazon Research Award.
May, 2026
I join Trustworthy Shopping AI Team at Amazon as an Applied Scientist intern at Seattle.
January, 2026
Our papers on Scalable Attack and Molecular Generation are accepted by ICLR 2026. See you in Rio de Janeiro!
May, 2025
I join Automated Reasoning Group at AWS AI as an Applied Scientist intern at New York.
May, 2025
Our paper "Modulated Diffusion" is accepted by ICML 2025. See you in Vancouver!
March, 2025
Our paper on "Scientific Discovery in Bioinformatics" is accepted by ICLR 2025 Workshop XAI4Science.
February, 2025
Our paper on Robustness Reprogramming is selected as a Spotlight paper (1.4% ≈ 162/11670) by ICLR 2025.
January, 2025
Our paper on Robustness Reprogramming is accepted by ICLR 2025. See you in Singapore!
October, 2024
I give a tutorial at DSAA about "Adversarial Robustness in Graph Neural Networks".
September, 2024
Our papers on Robust Transformers, Robust GNNs, and Efficient Certifiable Robustness are accepted by NeurIPS 2024. See you in Vancouver!
July, 2024
I receive the National AI Research Resource Pilot Award for our research on exploring and enhancing the robustness of LLMs and foundation models.
May, 2024
I join Amazon AWS GuardDuty as an Applied Scientist intern at New York.
May, 2024
Our tutorial on "Adversarial Robustness in Graph Neural Networks" is accepted by DSAA 2024. See you in San Diego!
May, 2024
I receive the Summer Graduate Merit Awards.
March, 2024
Our paper on Robust Transformers is accepted by ICLR 2024 Workshop on Reliable and Responsible Foundation Models.
February, 2024
Our paper on Single-Cell ATAC-seq Analysis is accepted by Briefings in Bioinformatics.
September, 2023
Our paper on Physical Dynamics is accepted by NeurIPS 2023. See you in New Orleans!
September, 2023
One paper on Explainable Survival Analysis is accepted by BMC Bioinformatics.
June, 2023
I join Big Search team in Baidu, Inc. as a research intern at Beijing.
March, 2022
I join Institute for AI Industry Research, Tsinghua University as a research intern.
Selected Publications
Full Publications →
ICLR 2026
Paper ↗
Fine-Grained Iterative Adversarial Attacks with Limited Computation Budget
ProTransformer: Robustify Transformers via Plug-and-Play Paradigm
Equivariant Spatio-Temporal Attentive Graph Networks to Simulate Physical Dynamics
Preprint
Paper ↗
Boosting Adversarial Robustness and Generalization with Structural Prior
Preprint
Paper ↗
Towards One-for-All Robustness Across a Continuum of Threat Levels
Internship Experience
May 2026
—
Aug 2026
Applied Scientist Intern
Amazon Web ServicesTrustworthy Shopping AI Team SEATTLE, WA
Collaborators: Rui Song, Lingdao Shao, Xueyu Mao, Yang Liu, Peijie Qiu
May 2025
—
Aug 2025
Applied Scientist Intern
Amazon Web ServicesAutomated Reasoning at AWS AI NEW YORK, NY
Collaborators: Ferhat Erata, MohamadAli Torkamani
May 2024
—
Aug 2024
Applied Scientist Intern
Amazon Web ServicesSecurity Analytics and AI Research (SAAR) NEW YORK, NY
Collaborator: MohamadAli Torkamani
May 2023
—
Aug 2023
Mar 2022
—
Sep 2022
Research Intern
Tsinghua UniversityInstitute for AI Industry Research BEIJING, CHINA
Collaborators: Wenbing Huang, Yu Rong