Publications

2026

  1. The Trap of Trajectory: Towards Understanding and Mitigating Spurious Correlations in Agentic Memory
    Luoxi Tang, Rupali Rajendra Vaje, Yuqiao Meng, and 5 more authors
    arXiv preprint arXiv:2605.09330, 2026
  2. EquiMem: Calibrating Shared Memory in Multi-Agent Debate via Game-Theoretic Equilibrium
    Yuqiao Meng, Sakshi Sunil Narvekar, Luoxi Tang, and 4 more authors
    arXiv preprint arXiv:2605.09278, 2026
  3. ER-MIA: Black-Box Adversarial Memory Injection Attacks on Long-Term Memory-Augmented Large Language Models
    Mitchell Piehl, Zhaohan Xi, Zuobin Xiong, and 2 more authors
    arXiv preprint arXiv:2602.15344, 2026
  4. TheraMind: a multi-LLM ensemble for accelerating drug repurposing in lung cancer via case report mining
    Vrushket More, Lyra Lu, Zeyu Ding, and 3 more authors
    npj Precision Oncology, 2026
    HTML
  5. The value of variance: Mitigating debate collapse in multi-agent systems via uncertainty-driven policy optimization
    Luoxi Tang, Yuqiao Meng, Joseph Costa, and 3 more authors
    In Proceedings of the 43rd International Conference on Machine Learning, 2026
    HTML
  6. Small agent group is the future of digital health
    Yuqiao Meng, Luoxi Tang, Dazheng Zhang, and 5 more authors
    In Proceedings of the 43rd International Conference on Machine Learning, 2026
    HTML
  7. Benchmarking LLM-Assisted Blue Teaming via Standardized Threat Hunting
    Yuqiao Meng, Luoxi Tang, Feiyang Yu, and 4 more authors
    In Proceedings of the 43rd International Conference on Machine Learning, 2026
  8. On the eligibility of LLMs for counterfactual reasoning: a decompositional study
    Shuai Yang, Qi Yang, Luoxi Tang, and 4 more authors
    In The Fourteenth International Conference on Learning Representations, 2026
    HTML

2025

  1. Data to Defense: The Role of Curation in Aligning Large Language Models Against Safety Compromise
    Xiaoqun Liu, Jiacheng Liang, Luoxi Tang, and 3 more authors
    In Proceedings of the 2025 Conference on Empirical Methods in Natural Language Processing, 2025
  2. Are LLMs Ready for English Standardized Tests? A Benchmarking and Elicitation Perspective
    Luoxi Tang, Tharunya Sundar, Shuai Yang, and 8 more authors
    arXiv preprint arXiv:2505.17056, 2025
  3. Uncovering vulnerabilities of llm-assisted cyber threat intelligence
    Yuqiao Meng, Luoxi Tang, Feiyang Yu, and 4 more authors
    arXiv preprint arXiv:2509.23573, 2025
  4. POLAR: Automating Cyber Threat Prioritization through LLM-Powered Assessment
    Luoxi Tang, Yuqiao Meng, Ankita Patra, and 3 more authors
    arXiv preprint arXiv:2510.01552, 2025

2024

  1. Zodiac: A cardiologist-level llm framework for multi-agent diagnostics
    Yuan Zhou, Peng Zhang, Mengya Song, and 5 more authors
    arXiv preprint arXiv:2410.02026, 2024
  2. On the difficulty of defending contrastive learning against backdoor attacks
    Changjiang Li, Ren Pang, Bochuan Cao, and 4 more authors
    In 33rd USENIX Security Symposium (USENIX Security 24), 2024

2023

  1. Defending pre-trained language models as few-shot learners against backdoor attacks
    Zhaohan Xi, Tianyu Du, Changjiang Li, and 5 more authors
    Advances in Neural Information Processing Systems, 2023
  2. On the security risks of knowledge graph reasoning
    Zhaohan Xi, Tianyu Du, Changjiang Li, and 6 more authors
    In 32nd USENIX Security Symposium (USENIX Security 23), 2023
  3. The dark side of automl: Towards architectural backdoor search
    Ren Pang, Changjiang Li, Zhaohan Xi, and 2 more authors
    In The 11th International Conference on Learning Representations, 2023
  4. An embarrassingly simple backdoor attack on self-supervised learning
    Changjiang Li, Ren Pang, Zhaohan Xi, and 4 more authors
    In Proceedings of the IEEE/CVF International Conference on Computer Vision, 2023

2022

  1. Trojanzoo: Towards unified, holistic, and practical evaluation of neural backdoors
    Ren Pang, Zheng Zhang, Xiangshan Gao, and 5 more authors
    In 2022 IEEE 7th European Symposium on Security and Privacy (EuroS&P), 2022
  2. Seeing is living? rethinking the security of facial liveness verification in the deepfake era
    Changjiang Li, Li Wang, Shouling Ji, and 4 more authors
    In 31st USENIX Security Symposium (USENIX Security 22), 2022
  3. On the security risks of {AutoML}
    Ren Pang, Zhaohan Xi, Shouling Ji, and 2 more authors
    In 31st USENIX Security Symposium (USENIX Security 22), 2022

2021

  1. Graph backdoor
    Zhaohan Xi, Ren Pang, Shouling Ji, and 1 more author
    In 30th USENIX security symposium (USENIX Security 21), 2021