Hello, I am Jeong Min Lee, a Research Scientist at Meta working on large-scale personalization and ranking systems. My work has focused on learning robust objectives from noisy implicit feedback, aligning optimization targets with longer-horizon user value, and improving decision systems under bias and uncertainty at internet scale.
I am currently especially interested in post-training for reasoning and agentic systems, including reward modeling, evaluation, memory, and iterative self-improvement.
Previously, I received my Ph.D. in Computer Science from the University of Pittsburgh, advised by Dr. Milos Hauskrecht. Across both academic and industry settings, a recurring theme in my work has been building adaptive learning systems from noisy, heterogeneous signals.
Current interests: post-training for reasoning and agents; personalization and memory; reward/preference modeling from noisy feedback; alignment to long-horizon value; scalable evaluation of learning systems.