Wonjeong Choi
Ph.D. Candidate, KAIST EE
Research Interest: Post-Training for Large Models, OOD Robustness, Trustworthy AI
E-mail: dnjswjd5457 (at) kaist.ac.kr
I am a Ph.D. student in School of Electrical Engineering (EE), KAIST, under the supervision of Prof. Jaekyun Moon (in MoonLab). Our laboratory has been studying in a wide range of research topics on hardware-friendly and robust machine learning algorithms. My own research emphasizes on developing trustworthy AI systems for safety-critical applications such as automotive, medical diagnosis, and manufactorying systems. In particular, I have explored the following areas:
1) Post-Training for Large Models: Test-Time Adaptation, Mask Tuning, Few-Shot Tuning, Transfer Learning
2) Out-of-Domain (OOD) Robustness: Domain Adaptation/Generalization
3) Trustworthy AI: LLM Safety, Model Calibration, Adversarial Attack