Shanshan Han

About

I'm a Research Engineer at ByteDance. My research focuses on building trustworthy AI systems, addressing safety and security challenges across the entire AI lifecycle.

Keywords: AI Safety Data Security LLM Guardrails FL Security Data Management

Research Overview

My research addresses AI safety and security across multiple layers of AI systems:

  1. Infrastructure Layer. Secure data foundations [PDF]
  2. Training Layer. Robust federated learning [PDF1] [PDF2]
  3. Inference Layer. Trustworthy LLM inference and LLM guardrail pipelines [PDF]
  4. Application Layer. Access control for retrieval-augmented generation systems [PDF]
Research overview across the AI lifecycle
Safety and security across AI lifecycles.

Selected Works

  1. FedSecurity: A Benchmark for Attacks and Defenses in Federated Learning and Federated LLMs
    Shanshan Han, Baturalp Buyukates, Zijian Hu, Han Jin, Weizhao Jin, Lichao Sun, Xiaoyang Wang, Wenxuan Wu, Chulin Xie, Yuhang Yao, Kai Zhang, Qifan Zhang, Yuhui Zhang, Salman Avestimehr, Chaoyang He
    Proceedings of the 30th ACM SIGKDD Conference on Knowledge Discovery and Data Mining (KDD 2024)
    [PDF] Invited Talk @ AI TIME
  2. Kick Bad Guys Out! Conditionally Activated Anomaly Detection in Federated Learning with Zero-Knowledge Proof Verification
    Shanshan Han, Wenxuan Wu, Baturalp Buyukates, Weizhao Jin, Qifan Zhang, Yuhang Yao, Salman Avestimehr
    NDSS-PRISM 2026
    [PDF]
  3. Veil: Storage and Communication Efficient Volume Hiding Algorithms
    Shanshan Han, Vishal Chakraborty, Michael Goodrich, Sharad Mehrotra, Shantanu Sharma
    Proceedings of the ACM on Management of Data (SIGMOD 2024)
    [PDF] Invited Talk @ Cryptography Group, MongoDB Inc.
  4. An Iterative Scheme for Leverage-based Approximate Aggregation
    Shanshan Han, Hongzhi Wang, Jialin Wan, Jianzhong Li
    IEEE 35th International Conference on Data Engineering (ICDE 2019)
    [PDF]
  5. Don't Be a Pot Stirrer! Authorized Vector Data Retrieval via Access-Aware Indexing
    Shanshan Han, Vishal Chakraborty, Sharad Mehrotra
    [PDF]
  6. Bridging the Safety Gap: A Guardrail Pipeline for Trustworthy LLM Inferences
    Shanshan Han, Salman Avestimehr, Chaoyang He
    [PDF] Invited Talk @ Ploutos AI Community
  7. FedML-HE: An Efficient Homomorphic-Encryption-Based Privacy-Preserving Federated Learning System
    Weizhao Jin, Yuhang Yao, Shanshan Han, Carlee Joe-Wong, Srivatsan Ravi, Salman Avestimehr, Chaoyang He
    FL@FM-NeurIPS 2023 Workshop
    [PDF]
  8. Fox-1: Open Small Language Model For Cloud And Edge
    Zijian Hu, Jipeng Zhang, Rui Pan, Zhaozhuo Xu, Shanshan Han, Han Jin, Alay Dilipbhai Shah, Dimitris Stripelis, Yuhang Yao, Salman Avestimehr, Chaoyang He, Tong Zhang
    [Technical Report]
  9. Alopex: A Computational Framework for Enabling On-Device Function Calls with LLMs
    Yide Ran, Zhaozhuo Xu, Yuhang Yao, Zijian Hu, Shanshan Han, Han Jin, Alay Dilipbhai Shah, Jipeng Zhang, Dimitris Stripelis, Tong Zhang, Salman Avestimehr, Chaoyang He
    [PDF]

Vision Papers

  1. Bridging Today and the Future of Humanity: AI Safety in 2024 and Beyond
    Shanshan Han
    [PDF]
  2. LLM Multi-Agent Systems: Challenges and Open Problems
    Shanshan Han, Qifan Zhang, Weizhao Jin, Zhaozhuo Xu
    [PDF]

Thesis

Safeguarding AI Lifecycles in the Cloud: Secure Data Management for Data at Rest, in Transit, and in Use
Ph.D. Dissertation