Wenjie Qu Email: wenjiequ@berkeley.edu Google Scholar

About Me

Hi! I am a PostDoc at UC Berkeley with Prof. Dawn Song. I received my PhD from the National University of Singapore.

My research focuses on the safety and security of LLMs and LLM agents. Previously, I worked on making AI systems verifiable and auditable through cryptography and watermarking.

I am always open to collaborations. If you share similar interests, feel free to email me.

Education

• PhD, National University of Singapore, 2023 – 2026
  Defense committee: Jiaheng Zhang (advisor), Bingsheng He, Xiaokui Xiao, Prateek Saxena

• B.E., Huazhong University of Science and Technology, 2019 – 2023

Publications (* equal contribution)

I have published 8 first-author papers at the top-tier security and cryptography conferences (IEEE S&P, USENIX Security, CCS, NDSS, CRYPTO).

Show all publications

• MetaPI: Constructing Prompt Injection Benchmarks from Any Agent Benchmarks [paper] Agent Security
  Peiran Wang, Chong Xiang, Wenjie Qu, Ying Li, Jiaheng Zhang, G. Edward Suh, Yuan Tian
  NeurIPS, 2026

• Understanding and Evaluating Claw-like Agent Security Through a Computer-Systems Lens [paper] [code] [website] [media coverage (Chinese)] Agent Security
  Peizhi Niu*, Wenjie Qu*, Shangding Gu, Tianneng Shi, Yuankai Li, Ahmad Tawaha, Hend Alzahrani, Vincent Siu, Boyi Li, Chenguang Wang, Jiaheng Zhang, Basel Alomair, Ming Jin, Muhao Chen, Chi Wang, Costas Spanos, Dawn Song
  arXiv, 2026

• Securing LLM Agents Need Intent-to-Execution Integrity [paper] Agent Security
  Wenjie Qu, Ming Xu, Peiran Wang, Shengfang Zhai, Jiaheng Zhang, Dawn Song
  arXiv, 2026

• AliMark: Enhancing Robustness of Sentence-Level Watermarking Against Text Paraphrasing [paper] [code] LLM Security
  Yuexin Li*, Wenjie Qu*, Linyu Wu, Yulin Chen, Yufei He, Tri Cao, Bryan Hooi, Jiaheng Zhang
  ICML, 2026

• Silent Leaks: Implicit Knowledge Extraction Attack on RAG Systems [paper] [code] [media coverage (Chinese)] LLM Security
  Yuhao Wang*, Wenjie Qu*, Shengfang Zhai, Yanze Jiang, Zichen Liu, Yue Liu, Yinpeng Dong, Jiaheng Zhang
  ICLR, 2026

• Self-Sovereign Agent [paper] [website] [media coverage] Agent Security
  Wenjie Qu, Xuandong Zhao, Jiaheng Zhang, Dawn Song
  arXiv, 2026

• IMMACULATE: A Practical LLM Auditing Framework via Verifiable Computation [paper] [code] [media coverage (Chinese)] Verifiable AICryptography
  Yanpei Guo, Wenjie Qu, Linyu Wu, Shengfang Zhai, Lionel Z. Wang, Ming Xu, Yue Liu, Binhang Yuan, Dawn Song, Jiaheng Zhang
  arXiv, 2026

• VerfCNN, Optimal Complexity zkSNARK for Convolutional Neural Networks [paper] [code] Verifiable AICryptography
  Wenjie Qu*, Yanpei Guo*, Yue Ying, Jiaheng Zhang
  IEEE S&P, 2026

• UltraProofs: Scalable Reed-Solomon Code Commitment [paper] Cryptography
  Yanpei Guo, Alex Xiong, Wenjie Qu, Jiaheng Zhang
  IEEE S&P, 2026

• TensorZKP: Repurposing GPU Tensor Cores for High-Performance Zero-Knowledge Proofs [paper] Cryptography
  Tao Lu, Jipeng Zhang, Yanpei Guo, Xuanming Liu, Wenjie Qu, Zonghui Wang, Wenzhi Chen, Jiaheng Zhang
  USENIX Security, 2026

• Rarus: A Succinct and Efficient Range Proof for Polynomial-based Vector Commitment [paper] Cryptography
  Xinyang Yang*, Wenjie Qu*, Yanpei Guo, Jiaheng Zhang
  USENIX Security, 2026

• Efficient SNARK for Floating-Point Computation Cryptography
  Wenjie Qu, Zihao Zhang, Jiaheng Zhang
  CCS, 2026

• Zero-Knowledge Proofs for Gradient Boosted Decision Trees Verifiable AICryptography
  Jiacheng Gao, Wenjie Qu, Yuan Zhang, Sheng Zhong, Jiaheng Zhang
  CCS, 2026

• Melior, Verifiable Mixed-Integer Programming via zkSNARKs Cryptography
  Wenjie Qu*, Yanze Jiang*, Yanpei Guo, Jiaheng Zhang
  CCS, 2026

• Celer: A Lookup Argument for Large-Scale Queries [paper] [code] Cryptography
  Wenjie Qu*, Yanpei Guo*, Zhen Xuan, Xuanming Liu, Jiaheng Zhang
  CRYPTO, 2026

• V3DB: Audit-on-Demand Zero-Knowledge Proofs for Verifiable Vector Search over Committed Snapshots [paper] [code] Verifiable AICryptography
  Zipeng Qiu*, Wenjie Qu*, Jiaheng Zhang, Binhang Yuan
  VLDB, 2026

• Prompt Inversion Attack against Collaborative Inference of Large Language Models [paper] [code] LLM Security
  Wenjie Qu, Yuguang Zhou, Yongji Wu, Tingsong Xiao, Binhang Yuan, Yiming Li, Jiaheng Zhang
  IEEE S&P, 2025

• HyperPianist: Pianist with Linear-Time Prover and Logarithmic Communication Cost [paper] Cryptography
  Chongrong Li, Pengfei Zhu, Yun Li, Cheng Hong, Wenjie Qu, Jiaheng Zhang
  IEEE S&P, 2025

• Provably Robust Multi-bit Watermarking for AI-generated Text [paper] [code] LLM Security
  Wenjie Qu, Wengrui Zheng, Tianyang Tao, Dong Yin, Yanze Jiang, Zhihua Tian, Wei Zou, Jinyuan Jia, Jiaheng Zhang
  USENIX Security, 2025

• zkGPT: An Efficient Non-interactive Zero-knowledge Proof Framework for LLM Inference [paper] [code] Verifiable AICryptography
  Wenjie Qu, Yijun Sun, Xuanming Liu, Tao Lu, Yanpei Guo, Kai Chen, Jiaheng Zhang
  USENIX Security, 2025

• DeepFold: Efficient Multilinear Polynomial Commitment from Reed-Solomon Code and Its Application to Zero-knowledge Proofs [paper] Cryptography
  Yanpei Guo, Xuanming Liu, Kexi Huang, Wenjie Qu, Tianyang Tao, Jiaheng Zhang
  USENIX Security, 2025

• Efficient Input-level Backdoor Defense on Text-to-Image Synthesis via Neuron Activation Variation [paper] [code] AI Security
  Shengfang Zhai, Jiajun Li, Yue Liu, Huanran Chen, Zhihua Tian, Wenjie Qu, Qingni Shen, Ruoxi Jia, Yinpeng Dong, Jiaheng Zhang
  ICCV, 2025, Highlight

• Reliable and Responsible Foundation Models [paper] LLM Security
  Xinyu Yang*, Junlin Han*, Rishi Bommasani*, Jinqi Luo*, Wenjie Qu*, Wangchunshu Zhou*, et al.
  TMLR, 2025

• PANDORA: Detailed LLM Jailbreaking via Collaborated Phishing Agents with Decomposed Reasoning [paper] LLM Security
  Zhaorun Chen, Zhuokai Zhao, Wenjie Qu, Zichen Wen, Zhiguang Han, Zhihong Zhu, Jiaheng Zhang, Huaxiu Yao
  ICLR Workshop on Secure and Trustworthy LLMs, 2024

• REaaS: Enabling Adversarially Robust Downstream Classifiers via Robust Encoder as a Service [paper] AI Security
  Wenjie Qu, Jinyuan Jia, Neil Gong
  NDSS, 2023

• MPass: Bypassing Learning-based Static Malware Detectors [paper] AI for Cybersecurity
  Jialai Wang, Wenjie Qu, Yi Rong, Han Qiu, Qi Li, Zongpeng Li, Chao Zhang
  DAC, 2023

• MultiGuard: Provably Robust Multi-label Classification against Adversarial Examples [paper] [code] AI Security
  Jinyuan Jia*, Wenjie Qu*, Neil Gong
  NeurIPS, 2022, Spotlight

• jTrans: Jump-Aware Transformer for Binary Code Similarity Detection [paper] [code] AI for Cybersecurity
  Hao Wang*, Wenjie Qu*, Gilad Katz, Wenyu Zhu, Zeyu Gao, Han Qiu, Jianwei Zhuge, Chao Zhang
  ISSTA, 2022

• EncoderMI: Membership Inference against Pre-trained Encoders in Contrastive Learning [paper] AI Security
  Hongbin Liu*, Jinyuan Jia*, Wenjie Qu, Neil Gong
  CCS, 2021

Academic Service

• Program Committee: CCS (2025, 2026), USENIX Security 2027, SaTML 2027

• Journal Reviewer: IoT-J, TIFS

• Reviewer: ICML 2022, MM 2024, COLM 2025, NeurIPS 2025

• Sub-Reviewer: PKC 2024

Selected Awards and Honors

• 2026 ML and Systems Rising Stars (39 recipients globally, one of only two from Asian institutions)

• 2026 Apple AI/ML PhD Fellowship Nomination (top 3 in NUS)

• 2025 Dean's Graduate Research Excellence Award (top 7 in department)

• 2025 DAAD AInet Fellowship

• 2025 Ant Group InTech Scholarship Future (10 recipients globally)

• 2023 NUS President's Graduate Fellowship

• 2022 Huawei Scholarship (sole recipient in department)

• 2020 National Scholarship

• 2018 National Olympiad in Informatics Winter Camp, Bronze Medal

Misc

Outside of research, I enjoy playing chess, a variety of board games, bridge, working out, and reading novels.