Biography
Jiaqi Xue is a fourth-year Ph.D. candidate in Computer Science at the University of Central Florida, advised by
Prof. Qian Lou. He builds efficient and secure AI systems, spanning efficient LLM serving and routing, secure
and trustworthy AI, privacy-preserving machine learning, and agentic code generation.
His recent work focuses on efficient LLM serving and routing: R2-Router (ICML 2026) lets a router reason about
which model to use and how much computation to spend, and was the first method to push the RouterArena
leaderboard beyond a score of 70; Chain-of-Route (NeurIPS 2026) extends routing from single queries to
stateful, multi-turn agentic interactions; and HW‑Router (DAC 2026) makes routing hardware-aware for
scalable multi-LLM serving. In secure and trustworthy AI, TrojLLM (NeurIPS 2023) introduced the first
automated prompt-injected Trojan attack on large language models, and PRO (NeurIPS 2026) enables precise and
robust watermarks for open-source LLMs. In privacy-preserving machine learning, DictPFL (NeurIPS 2025) makes
federated learning on FHE-encrypted gradients efficient, and CipherPrune (ICLR 2025) scales hybrid
MPC–FHE private Transformer inference. Most recently, he has been working on agentic code generation,
from FHE-Coder (ICLR 2026) to multi-agent protocols and harnesses at Amazon Kiro Science.
To date, he has published 17 papers in leading venues across machine learning (NeurIPS, ICML, ICLR),
NLP and vision (ACL, EMNLP, NAACL, ECCV), security and privacy (IEEE S&P, PoPETs, SaTML), and systems
(DAC, PACT). He is the first or co-first author of nine of them: four at NeurIPS and others at ICML, ECCV,
EMNLP, etc. He has interned at Samsung Research America and Amazon Kiro Science.
Experience
-
Amazon Kiro Science, Santa Clara, CA
Applied Scientist Intern, May 2026 - Aug. 2026, focused on multi-agent system (MAS) protocols and
harnesses for code generation
-
Samsung Research America, Mountain View, CA
AI Research Intern, May 2024 - Aug. 2024, focused on LLM security
-
University of Central Florida, Orlando, FL
Graduate Research Assistant, Jan. 2023 - Present
Publications
(* indicates equal contribution)
-
Chain-of-Route: State-Aware LLM Routing for Multi-Turn Conversations
Jiaqi Xue, Mengxin Zheng, Heng Huang, Qian Lou
Annual Conference on Neural Information Processing Systems (NeurIPS), 2026
-
PRO: Enabling Precise and Robust Text Watermark for
Open-Source LLMs
Jiaqi Xue, Yifei Zhao, Mansour Al Ghanim, Shangqian Gao, Ruimin Sun, Qian Lou,
Mengxin Zheng
Annual Conference on Neural Information Processing Systems (NeurIPS), 2026
-
R2-Router: A New Paradigm for LLM Routing with
Reasoning
Jiaqi Xue, Qian Lou, Jiarong Xing, Heng Huang
International Conference on Machine Learning (ICML), 2026
First method to push the RouterArena leaderboard beyond a score of 70
-
FHE-Coder: Secure Agentic Code Generation for Fully
Homomorphic Encryption
Mayank Kumar, Jiaqi Xue, Mengxin Zheng, Qian Lou
International Conference on Learning Representations (ICLR), 2026
-
HW-Router: Hardware-Aware Routing for Scalable
Multi-LLM Serving
Ahasan Kabir, Jiaqi Xue, Mengxin Zheng, Qian Lou
Design Automation Conference (DAC), 2026
-
SoK: Can Fully Homomorphic Encryption Support General AI
Computation? A Functional and Cost Analysis
Jiaqi Xue, Xin Xin, Wei Zhang, Mengxin Zheng, Qianqian Song, Minxuan Zhou,
Yushun Dong, Dongjie Wang, Xun Chen, Jiafeng Xie, Liqiang Wang, David Mohaisen, Hongyi Wu, Qian Lou
Proceedings on Privacy Enhancing Technologies (PoPETs), 2026
-
RobPI: Robust Private Inference against Malicious Client
Jiaqi Xue, Mengxin Zheng, Qian Lou
IEEE Conference on Secure and Trustworthy Machine Learning (SaTML), 2026
-
DictPFL: Efficient and Private Federated Learning on
Encrypted Gradients
Jiaqi Xue, Mayank Kumar, Yuzhang Shang, Shangqian Gao, Rui Ning, Mengxin Zheng,
Xiaoqian Jiang, Qian Lou
Annual Conference on Neural Information Processing Systems (NeurIPS), 2025
-
Evaluating the Robustness and Accuracy of Text Watermarking
Under Real-World Cross-Lingual Manipulations
Mansour Al Ghanim, Jiaqi Xue, Rochana Prih Hastuti, Mengxin Zheng, Yan Solihin,
Qian Lou
Conference on Empirical Methods in Natural Language Processing (EMNLP), 2025
-
CipherPrune: Efficient and Scalable Private Transformer
Inference
Yancheng Zhang, Jiaqi Xue, Mengxin Zheng, Mimi Xie, Mingzhe Zhang, Lei Jiang,
Qian Lou
International Conference on Learning Representations (ICLR), 2025
-
DataSeal:
Ensuring the Verifiability of Private Computation on Encrypted Data
Muhammad Husni Santriaji, Jiaqi Xue, Yancheng Zhang, Qian Lou, Yan Solihin
IEEE Symposium on Security and Privacy (IEEE S&P), 2025
-
BadFair: Backdoored Fairness Attacks with Group-conditioned
Triggers
Jiaqi Xue, Qian Lou, Mengxin Zheng
Conference on Empirical Methods in Natural Language Processing (EMNLP), 2024
-
SSL-Cleanse: Trojan Detection and Mitigation in
Self-Supervised Learning
Mengxin Zheng*, Jiaqi Xue*, Zihao Wang, Xun Chen, Qian Lou, Lei Jiang, Xiaofeng
Wang
European Conference on Computer Vision (ECCV), 2024
-
CR-UTP: Certified Robustness against Universal Text
Perturbations on Large Language Models
Qian Lou, Xin Liang*, Jiaqi Xue*, Yancheng Zhang, Rui Xie, Mengxin Zheng
Annual Meeting of the Association for Computational Linguistics (ACL), 2024
-
TrojFSP: Trojan Insertion in Few-shot Prompt Tuning
(Oral)
Mengxin Zheng, Jiaqi Xue, Xun Chen, Yanshan Wang, Qian Lou, Lei Jiang
Annual Conference of the North American Chapter of the Association for Computational Linguistics
(NAACL), 2024
-
BoostCom: Towards Efficient Universal Fully Homomorphic
Encryption by Boosting the Word-wise Comparisons
Ardhi Wiratama Baskara Yudha, Jiaqi Xue, Qian Lou, Huiyang Zhou, Yan Solihin
International Conference on Parallel Architectures and Compilation Techniques (PACT),
2024
-
TrojFair: Trojan Fairness Attacks
Jiaqi Xue, Mengxin Zheng, Yi Sheng, Lei Yang, Qian Lou, Lei Jiang
Proceedings of the 1st ACM Workshop on Large AI Systems and Models with Privacy and Safety Analysis
(LAMPS), 2024
-
CryptoTrain: Fast Secure Training on Encrypted Dataset
Jiaqi Xue, Yancheng Zhang, Yanshan Wang, Xueqiang Wang, Hao Zheng, Qian Lou
Proceedings of the 1st ACM Workshop on Large AI Systems and Models with Privacy and Safety Analysis
(LAMPS), 2024
-
TrojLLM: A Black-box Trojan Prompt Attack on Large Language
Models
Jiaqi Xue, Mengxin Zheng, Ting Hua, Yilin Shen, Yepeng Liu, Ladislau Bölöni,
Qian Lou
Annual Conference on Neural Information Processing Systems (NeurIPS), 2023
Awards
-
DAC Young Fellow, 2026
-
UCF Faculty Cluster Initiative (FCI) Scholarship, 2025
-
NeurIPS Top Reviewer Award, 2024
-
NeurIPS Scholar Award, 2023
Services
Reviewer:
ICML, ICLR, NeurIPS, AAAI, IJCAI, TMLR, ACL, EMNLP, ICCV, CVPR
Teaching
Graduate Teaching Assistant, University of Central Florida
-
CIS3360 – Security in Computing, Spring 2026
-
CDA5106 – Advanced Computer Architecture, Fall 2023, Fall 2024, Fall 2025
-
CAP6614 – Current Topics in Machine Learning, Spring 2024, Spring 2025
-
CDA3103 – Computer Logic and Organization, Summer 2023
© 2026 Jiaqi Xue