Resume
PDF
Education
- M.S.E. in Computer Science, Johns Hopkins University, Expected May 2027
- B.S. in Computer Science and Information Engineering, National Taiwan University, Jun 2025
Experience
Math Tutor, Johns Hopkins University, Baltimore, MD, Jan 2026 – Present
- Tutored students in grades 3–7 in mathematics at Henderson-Hopkins, providing individualized guidance and adapting explanations to different learning levels and needs.
Research Intern, Washington University in St. Louis, St. Louis, MO, Jun 2024 – Aug 2024
- Led a first-author project developing an adaptive input sanitization framework for legacy medical systems using extended Berkeley Packet Filter (eBPF), enabling runtime deployment of kernel-level defenses without service downtime while achieving <19 ms patch loading time.
- Developed an automated lighting-control system for the Open Platform for Cyber-Physical Systems (OPCPS) by implementing a custom network-based interface connecting software with physical lighting hardware.
Research Intern, National Taiwan University, Taipei, Taiwan, Jan 2024 – Jun 2024
- Collaborated with PhD students on gait-analysis research, evaluating monocular depth-estimation models and integrating open-source computer-vision pipelines to explore 3D depth recovery from 2D video.
Projects
Modeling Social Pressure in Multi-Agent LLM Environments, May 2026
- Built a Python-based multi-agent LLM simulation and evaluation framework using model APIs and prompt engineering to analyze cooperative behavior across 5 LLM families and 3 experimental conditions.
CompareAI: User-Centered AI Tool Comparison Platform, May 2026
- Designed an AI tool comparison platform in Figma through contextual inquiry, participatory design, and usability testing; usability tasks, including tool comparison, achieved 75–100% completion in ~20 seconds.
I-CALM: Incentivizing Confidence-Aware Abstention for LLM Hallucination Mitigation, Mar 2026
- Engineered and evaluated a confidence-aware LLM framework for reducing unreliable answers using model APIs and prompt engineering; tested 5 LLM families across 3 QA datasets, reducing false-answer rate from 52.3% to 34.2% on GPT-5 mini/PopQA.
Full-Stack Meeting Coordination Platform, Jun 2024
- Implemented frontend and backend components for a meeting coordination platform using Python, JavaScript, HTML, CSS, and MongoDB, contributing to an integrated system that streamlined meeting coordination.
Publications
Haotian Zong, Binze Li, Yufei Long, Sinyin Chang, Jialong Wu, Gillian K Hadfield. I-CALM: Incentivizing Confidence-Aware Abstention for LLM Hallucination Mitigation. Epistemic Intelligence in Machine Learning Workshop, co-located with ICML 2026
Sinyin Chang, Ao Li, Evin Jaff, Yuanhaur Chang, Jinwen Wang, Ning Zhang, Hsu-Chun Hsiao. AdapSan: Adaptive Input Sanitization in Medical Systems with eBPF. ACM Workshop on Adaptive and Autonomous Cyber Defense, 2024
Ao Li, Sinyin Chang, Guorui Li, Yuanhaur Chang, Nathan Fisher, Thidapat Chantem. Software and Behavior Diversification for Swarm Robotics Systems. ACM Workshop on Moving Target Defense, 2023
Technical Skills
- Programming: Python, C/C++, JavaScript, SQL, HTML, CSS
- Technologies & Tools: React, MongoDB, Git, Docker, Figma
- AI/ML: LLMs, Prompt Engineering, Multi-Agent Systems, LLM APIs, Agent/Model Evaluation
Volunteer
- Designer, Graduate Association, NTU
- Student Volunteer, Office of International Affairs, NTU
- Student Volunteer, Parker Jewish Institute for Health Care and Rehabilitation, NY
- International Ambassador, HSNU
Languages
- Mandarin, English, Spanish
