Publications

Publications by categories in reversed chronological order.

2026

  1. arXiv
    efficient-surrogate-analysis.png
    Efficient LLM Adversarial Training via Low-Rank Defense and Circuit-Guided Surrogates
    Weiyi He*, Yuping Lin, Tang Jiliang, and Yue Xing
    arXiv preprint arXiv:2607.28959, 2026
    PDF
  2. arXiv
    aura-framework.png
    AURA: Adaptive Uncertainty-aware Refinement for LLM-as-a-Judge Auditing
    Zilong Zhang, Yi-Ting Hung, Weiyi He, Junxi Zhang, Lei Ding, and Chi-Kuang Yeh
    arXiv preprint arXiv:2606.19714, 2026
    PDF
  3. AISTATS
    aistats-generalization-gap.png
    Impact of Positional Encoding: Clean and Adversarial Rademacher Complexity for Transformers under In-Context Regression
    Weiyi He* and Yue Xing
    International Conference on Artificial Intelligence and Statistics (AISTATS), 2026
    PDF Poster

2025

  1. ACL
    privacy-memory-attack.png
    Unveiling privacy risks in llm agent memory
    Bo Wang, Weiyi He, Shenglai Zeng, Zhen Xiang, Yue Xing, Jiliang Tang, and Pengfei He
    In Proceedings of the 63rd Annual Meeting of the Association for Computational Linguistics (ACL), 2025
    PDF Code