I am an undergraduate student in Artificial Intelligence at Chung-Ang University
in Seoul, Korea. My research focuses on adaptive post-training for foundation models:
understanding how supervision and training should adapt to a model's current knowledge, capabilities, and behavior.
At Trillion Labs, I worked on foundation-model post-training, with a focus
on knowledge distillation and continual learning for large language models. I am currently interested in
self-distillation and efficient continual post-training.
My long-term goal is to build self-evolving AI systems that can continuously improve by leveraging their own
knowledge and behavior. My prior work was nominated for an outstanding paper award
at EMNLP.
@inproceedings{jung-etal-2025-todi,
title = "{T}o{D}i: Token-wise Distillation via Fine-Grained Divergence Control",
author = "Jung, Seongryong and Yoon, Suwan and Kim, DongGeon and Lee, Hwanhee",
booktitle = "Proceedings of the 2025 Conference on Empirical Methods in Natural Language Processing",
year = "2025",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2025.emnlp-main.409",
}