I am a Frontier Education AI Research Engineer at Junyi Academy Foundation, where I lead research on AI safety evaluation and guardrail design for tutoring AI. My work asks how to measure whether AI assistance builds or erodes a learner’s independent capability, and what evidence regulators and auditors could require and verify.

My current research focuses on the cognitive risks AI-mediated learning poses to human agency — how everyday reliance on AI tools shapes a learner’s capacity for independent thought, and where the line falls between assistance that builds capability and assistance that erodes it. I’m working to frame this as a measurable technical problem in educational AI, connecting near-term dependency risk to the broader concern of gradual disempowerment as AI systems become more deeply embedded in how people learn and think.

I am an undergraduate in Life Sciences and Genome Sciences at National Yang Ming Chiao Tung University. I organize the technical AI safety research group at the National Taiwan University AI Safety Group (NTUAIS) and lead its post-training and AI control group, which works on reward hacking, emergent misalignment, agent evaluation and sandbagging. I also work as a research assistant in Dr. Jie-Fan Chang’s lab at National Taiwan University on machine learning for medical image analysis.

Before turning to safety, I built scalable, efficient AI engines for Edge AI product lines, including performance tuning, cross-platform deployment, and integration within AIoT systems. My research background spans neuroimaging, artificial intelligence and signal processing. I’ve developed deep learning–based solutions for medical image processing, emphasizing signal clustering, model robustness, and data augmentation. In studies of glucose metabolism in neurodegenerative disease, I designed a novel analysis framework to offer a new perspective on understanding cerebral metabolic connectivity — helping to interpret how diseases affect regional interactions in the brain.

Beyond research, I actively engage in digital rights and civic tech. I currently serve on the Taipei City Government Youth Advisory Committee, advising on AI usage regulation for public sectors, and contribute to the Judicial Reform Foundation’s Digital Law Taskforce, where I work on algorithmic transparency, automated decision-making transparency, and data sovereignty. I also collaborate with the Open Culture Foundation and g0v.tw, supporting public-interest technology and open culture initiatives.

💡 Recent thoughts at /now
(last updated: 2026-09-23)

---

Current Roles

• Frontier Education AI Research Engineer
  Junyi Academy Foundation
  Leading research and technical deployment aimed at mitigating AI-driven cognitive risks and gradual disempowerment in educational environments.
  

  Jul 2026 - Present
• Organizer
  National Taiwan University AI Safety Group (NTUAIS)
  Lead of Post-training and AI Control Research Group, focusing on reward hacking, emergent misalignment.
  

  Sep 2026 - Present
• Tech Contributor
  Judicial Reform Foundation
  Digital Law Task Force
  Collaborated work: legislation of digital bill of rights
  

  Oct 2024 – Present
• Research Assistant
  Department of Computer Science & Information Engineering, National Taiwan University
  Advanced AI research for medical image analysis, specializing in neuroimaging scans.
  XAI method design & research for clinical data interpretation.
  Developed robust medical image segmentation approach for clinical edge computing.
  

  Sep 2022 – Present