• Unlearn to Relearn Backdoors: Deferred Backdoor Functionality Attacks on Deep Learning Models arXiv (2024)

  Investigates deferred backdoor functionality attacks that survive machine unlearning and parameter scrubbing in deep neural networks.
• Mask-based Invisible Backdoor Attacks on Object Detection IEEE ICIP (2024)

  Proposes imperceptible mask-based backdoor insertion techniques tailored for object detection architectures.

• Mar 2026 - Mar 2028
  Korea University (고려대학교)
  M.S. in Artificial Intelligence

  Graduate research in Artificial Intelligence, focusing on Trustworthy AI, Agentic AI, and deep learning model security.

• Aug 2024 - Mar 2025
  POSTECH (포항공과대학교)
  AI Researcher

  Researched deep learning security and trustworthy AI mechanisms, focusing on backdoor attacks, model resilience, and representation defense.

• Agentcat과 반복 유입의 기술

  진부해 보이는 비즈니스 모델이어도 Threads에서 먹히는 훅을 찾아 반복 유입을 만드는 방식에 대한 메모.

Links

• github /
• linkedin /
• drifty /
• quotawake /
• workcat /
• email /
• [email protected]