Jeongjae Lee

I am a M.S. student at KAIST AI’s BISPL lab, advised by Dr. Jong Chul Ye. I received my B.S. in Electrical and Computer Engineering from Seoul National University (Summa Cum Laude).

Research Interests

My research interests broadly span denoising generative models and reinforcement learning. Over the past year, I have focused on RL-based fine-tuning of denoising generative models.

Recent Works

• [arXiv 2026] RSM: Introduced Reward Score Matching (RSM), a unifying view of reward-based fine-tuning for diffusion and flow-matching models that reduces a fragmented literature to a common objective and a smaller, more interpretable design space.

• [ICLR 2026] PCPO: Introduced Proportionate Credit Policy Optimization (PCPO), a policy-gradient objective for preference alignment of diffusion and flow-matching models that improves stability by (i) using a log-ratio surrogate that avoids exponentiation and (ii) correcting disproportionate credit assignment across timesteps.

Past Experience

• [Postbacc Intern @ BISPL] Worked on industrial co-op projects on medical AI.

• [Undergrad Intern @ DeepMetrics] Worked on automation of mechanical ventilator control in ICUs.

• [Undergrad Intern @ Steinegger Lab] Contributed to Foldseek, an open-source biotechnology software for protein structure comparison (published at Nature Biotechnology). Optimized LDDT alignment score computation, achieving a 13x speedup.