Profile Picture

Xi Li (李溪)

Assistant Professor, Ph.D.

TrustBrew Lab

Department of Computer Science

University of Alabama - Birmingham

Bio

I am an Assistant Professor in the Department of Computer Science at the University of Alabama at Birmingham. My research interests include Multimodal AI Safety, Adversarial Machine Learning, and Reliable AI Agents. More about my research can be found in the TrustBrew Lab. I received my Ph.D. and M.S. degrees in Computer Science from Penn State University, completing my doctoral studies under the guidance of Dr. George Kesidis and Dr. David Miller. I received my B.S. degree from the School of Information Science and Engineering at Southeast University.

Check my resume here (Last updated: Sep. 2026).

Publications

Benchmarking LLM-Assisted Blue Teaming via Standardized Threat Hunting

Yuqiao Meng, Luoxi Tang, Feiyang Yu, Xi Li, Guanhua Yan, Ping Yang, Zhaohan Xi

ICML 2026

Chain-of-Scrutiny: Detecting Backdoor Attacks for Large Language Models

Xi Li, Ruofan Mao, Yusen Zhang, Renze Lou, Chen Wu, Jiaqi Wang

ACL(Findings) 2025

Paper Code

PeerGuard: Defending Multi-Agent Systems Against Backdoor Attacks Through Mutual Reasoning

Falong Fan, Xi Li

Best Paper Award, IEEE IRI 2025

Paper

Securing Federated Learning Against Novel and Classic Backdoor Threats During Foundation Model Integration

Xiaohuan Bi, Xi Li

IEEE IRI 2025

Paper

AAAR-1.0: Assessing AI’s Potential to Assist Research

Renze Lou, Hanzi Xu, Sijia Wang, Jiangshu Du, Ryo Kamoi, Xiaoxin Lu, Jian Xie, Yuxuan Sun, Yusen Zhang, Jihyun Janice Ahn, Hongchao Fang, Zhuoyang Zou, Wenchao Ma, Xi Li, Kai Zhang, Congying Xia, Lifu Huang, Wenpeng Yin

ICML, 2025

Paper

Mitigating Image Captioning Hallucinations in Vision-Language Models

Fei Zhao, Chengcui Zhang, Runlin Zhang, Tianyang Wang, Xi Li

IEEE MIPR, 2025

Paper

Backdoor Inversion in Neural-Activation Space

Guangmingmei Yang, Xi Li, Hang Wang, David Miller, George Kesidis

IEEE MLSP, 2025

Foundation Models in Federated Learning: Assessing Backdoor Vulnerabilities

Xi Li, Chen Wu, Jiaqi Wang

IJCNN, 2025

Paper

Unveiling Backdoor Risks Brought by Foundation Models in Heterogeneous Federated Learning

Xi Li, Chen Wu, Jiaqi Wang

PAKDD, 2024

Paper Poster Code

Temporal-Distributed Backdoor Attack Against Video-Based Action Recognition

Xi Li, Songhe Wang, Ruiquan Huang, Mahanth Gowda, George Kesidis

AAAI, 2024

Paper Poster Code

A BIC-based Mixture Model Defense against Data Poisoning Attacks on Classifiers

Xi Li, David Miller, Zhen Xiang, George Kesidis

IEEE MLSP, 2023

Test-Time Detection of Backdoor Triggers of Poisoned Deep Neural Networks

Xi Li, David Miller, Zhen Xiang, George Kesidis

ICASSP, 2022

Paper Poster

Detecting Backdoor Attacks Against Point Cloud Classifiers

Zhen Xiang, David Miller, Siheng Chen, Xi Li, George Kesidis

ICASSP, 2022

Paper

A Backdoor Attack against 3D Point Cloud Classifiers

Zhen Xiang, David Miller, Siheng Chen, Xi Li, George Kesidis

ICCV, 2021

Paper

Adapting Vision Foundation Models with Cascaded Semantics

Xi Xiao, Xingjian Li, Cheng Han, Tianyang Wang, Lin Zhao, Yunbei Zhang, Guosheng Hu, Runmin Jiang, Xi Li, Xiao Wang, Min Xu

Transactions on Machine Learning Research (TMLR) 2026

Correcting the distribution of batch normalization signals for Trojan mitigation

Xi Li, Zhen Xiang, David Miller, George Kesidis

Neurocomputing, 2024

Paper

BIC-based Mixture Model Defense against Data Poisoning Attacks on Classifiers: A Comprehensive Study

Xi Li, David Miller, Zhen Xiang, George Kesidis

IEEE Transactions on Knowledge and Data Engineering (TKDE), 2024

Paper Code

Evolving Safety Landscape of Multi-modal Large Language Models: A Survey of Emerging Threats and Safeguards

Xi Li, Shu Zhao, Xiaohan Zou, Fei Zhao, Fuxiao Liu, Yusen Zhang, Cheng Han, Yushun Dong, Jiaqi Wang

ICLR Workshop on Principled Design for Trustworthy AI, 2026

Rethinking the Safety Landscape for Foundation Models: A Multi-Modal Perspective

Xi Li, Shu Zhao, Fei Zhao, Runlong Yu

ICCV 2025 T2FM Workshop

Paper

Position Paper: Assessing Robustness, Privacy, and Fairness in Federated Learning Integrated with Foundation Models

Xi Li, Jiaqi Wang

ICCV 2025 T2FM Workshop

Paper

Backdoor Threats from Compromised Foundation Models to Federated Learning

Xi Li, Songhe Wang, Chen Wu, Hao Zhou, Jiaqi Wang

FL@FM-NeurIPS'23

Paper Poster Code

Survey on Neural Network Parameter Acquisition: from Optimization to Generation

Jiaqi Wang, Yusen Zhang, Xi Li, Simeng Han, Zining Zhu, Lingjuan Lyu, Yaqing Wang, Ruiquan Huang, Arman Cohan, Rui Zhang, Sheng Li, Xin (Eric) Wang.

NeuroGen: Neural Network Parameter Generation via Large Language Models

Jiaqi Wang, Yusen Zhang, Xi Li

Under review

Paper

Grants & Awards

Active Research Projects

Awards

Teaching

Instructor UAB

Teaching Assistant PSU

Tutorials & Talks

Tutorials:

Talks:

Experience

Associate Scientist | 2025 - present

  • Integrative Center for Aging Research
  • The University of Alabama at Birmingham, School of Medicine
UAB Medicine Logo

Associate Scientist | 2025 - present

  • Center for the Study of Community Health
  • The University of Alabama at Birmingham, School of Public Health
UAB Public Health Logo

Machine Learning Engineer Intern | Summer 2024

  • Meta, NYC
  • Modern Recommendation System Team
  • Explored early fusion architecture for multimodal understanding.
Meta Logo

Service

Xi Li (李溪)
TrustBrew Lab