RESEARCH INTERESTS
- Trustworthy LLMs (Privacy & Security)
- NLP
- Federated Learning
EDUCATION
Duke University Ph.D. Student in Electrical and Computer Engineering Durham, NC Advisor: Prof. Yiran Chen | 09/2022 - PRESENT Durham, NC |
National Taiwan University(NTU) M.S. in Graduate Institute of Networking and Multimedia | 09/2017 - 06/2019 Taipei, Taiwan |
National Tsing Hua University(NTHU) B.S. in Interdisciplinary Program of Engineering Double Disciplines: Computer Science and Industrial Engineering & Engineering Management (Transfer from National Taiwan Ocean University to This Program in February 2014) | 02/2014 - 06/2017 Hsinchu, Taiwan |
Experience
Computational Evolutionary Intelligence Lab (Prof. Yiran Chen)
Graduate Research Assistant, Duke ECE
● Trustworthy Large Language Model
– Design and implementation of defense mechanisms to preserve users’ data privacy in LLM.
– Design and implementation of model alignment with safety policies while maintaining response fluency and coherence.
● Cognition and Neurosymbolic AI
– Design and implementation of pretraining methods for language models that integrates both syntax and semantic information.
● Federated Instruction Tuning
– Design and implementation of methods that explore federated learning based instruction tuning for LLMs
09/2022 - PRESENT
Accenture
Research Intern (Technology R&D Assoc Principal, Cyber Security Labs)
● Multi-shot jailbreaking attack and defense on Large Language Model
– Implementation of multi-shot jailbreaking attack methods and design of new defense strategies for large language models.
06/2024 - 08/2024
MediaTek Incorporated
Engineer, Intelligent Software Development
● The Owner of Android TV Netflix Audio
– Supporting of Android TV audio-related ICs and projects for Netflix certification
02/2020 - 06/2022
Microsoft Research Asia
Research Intern
● Question and Answering Language Model
– Design and implementation of a Question-Answering Language Model with contributions to
Bing’s Healthcare Question and Answering feature.
07/2018 - 09/2018
Machine Discovery and Social Network Mining Lab (Prof. Shou-De Lin)
Graduate Research Assistant, NTU Graduate Institute of Networking and Multimedia
● Food Recommendation
– Design and implementation of algorithms to improve the accuracy of food recommendation.
● Finance Prediction Project
– Design and implementation of natural language processing algorithms to improve the accuracy
of US treasury yield prediction.
● Decremental Learning Project
– Design and implementation of decremental Learning algorithm to improve classification
accuracy.
09/2017 - 06/2019
Project Participation at National Tsing Hua University
● MediaTek LinkIt One Prevent Leakage Project
– Design and implementation of a leakage-prevention project on MediaTek LinkIt One
– Finalist, 2016 Mobile Computer Workshop (Huisun Forest, Nantou, Taiwan)
05/2015 - 07/2015
Publications
- Martin Kuo, Jingyang Zhang, Jianyi Zhang, Minxue Tang, Louis DiValentin, Aolin Ding, Jingwei Sun, William Chen, Amin Hass, Tianlong Chen, Yiran Chen, and Hai Li. 2025b. Proactive privacy amnesia for large language models: Safeguarding PII with negligible impact on model utility. In The Thirteenth International Conference on Learning Representations (ICLR)
- Jingyang Zhang, Jingwei Sun, Eric Yeats, Yang Ouyang, Martin Kuo, Jianyi Zhang, Hao Yang, and Hai Li.
2025. Min-k%++: Improved baseline for pre-training data detection from large language models. In The
Thirteenth International Conference on Learning Representations (ICLR)
- Martin Kuo, Jianyi Zhang, Aolin Ding, Qinsi Wang, Louis DiValentin, Yujia Bao, Wei Wei, Da-Cheng Juan, Hai Li, and Yiran Chen. 2025a. H-cot: Hijacking the chain-of-thought safety reasoning mechanism to jailbreak large reasoning models, including openai o1/o3, deepseek-r1, and gemini 2.0 flash thinking
- Jianyi Zhang*, Saeed Vahidian*, Martin Kuo*, Chunyuan Li, Ruiyi Zhang, Tong Yu, Guoyin Wang, and Yiran Chen. 2024. Towards building the federatedgpt: Federated instruction tuning. In IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP), pages 6915–6919
- Martin Kuo, Jianyi Zhang, and Yiran Chen. 2023. Dacbert: Leveraging dependency agreement for cost-efficient bert pretraining. arXiv preprint arXiv:2311.04799
- Martin Kuo, Yaobo Liang, Lei Ji, Nan Duan, Linjun Shou, Ming Gong, and Peng Chen. 2020. Tag and correct: Question aware open information extraction with two-stage decoding. arXiv preprint arXiv:2009.07406
Li, and Yiran Chen. 2022. Next generation federated learning for edge devices: an overview. In 2022
IEEE 8th International Conference on Collaboration and Internet Computing (CIC), pages 10–15. IEEE
HONORS & AWARDS
2021
MediaTek Incorporated vAwarD
Mediatek Core Value Recognition “Conviction Inspired by Deep Thinking, Inclusiveness”
2020
MediaTek Incorporated vAwarD
Mediatek Core Value Recognition “Customer Focus, Constant Renewel”
MediaTek Incorporated vAwarD
Mediatek Core Value Recognition “Customer Focus”
2018
Microsoft Research Asia
Stars of Tomorrow Award (less than 10% of Interns at MSRA)