Lee, Jaeseong (이재성)
Ph.D. alum
Department of Computer Science and Engineering
Dissertation: Expert Subnetworks for Data- and Compute-Efficient Language Models
Research Interests
- LLM Inference Optimization
- Low-resource Languages & Multilingual
Education
- Ph.D. in Computer Science, Seoul National University, 2026
- M.S. in Computer Science, Seoul National University, 2021
- B.S. in Mathematical Education & Computer Science, Seoul National University, 2019
Publications
- TAGQuant: Token-Aware Clustering for Group-Wise Quantization, To appear at EACL 2026 (Industry Track)
Jaeseong Lee, Seung-won Hwang, Aurick Qiao, Zhewei Yao, Yuxiong He. - TALE: Token-Adaptive Low-Rank KVCache Approximation with Reconstruction Elimination, TACL 2025
Jaeseong Lee, Seung-won Hwang, Aurick Qiao, Daniel Campos, Zhewei Yao, Yuxiong He. - STUN: Structured-Then-Unstructured Pruning for Scalable MoE Pruning, ACL 2025
Jaeseong Lee, Seung-won Hwang, Aurick Qiao, Daniel Campos, Zhewei Yao, Yuxiong He. - Smarter, Not Harder: Training-Free Adaptive Computation for Transformers, Findings of ACL 2025
{Romain Storaï, Jaeseong Lee}, Seung-won Hwang. - PROM: Pivoted and Regulated Optimization for Multilingual Instruction Learning, NAACL 2025
Jaeseong Lee, Seung-won Hwang, Hojin Lee, Yunju Bak, Changmin Lee. - PLEX: Adaptive Parameter-Efficient Fine-Tuning for Code LLMs using Lottery-Tickets, NAACL 2025 (Industry Track)
{Jaeseong Lee, Hojae Han}, Jongyoon Kim, Seung-won Hwang, Naun Kang, KyungJun An, Sungho Jang. - Breaking ReLU Barrier: Generalized MoEfication for Dense Pretrained Models, EMNLP 2024
Jaeseong Lee, Seung-won Hwang, Wonpyo Park, Mingi Ji. - Interventional Speech Noise Injection for ASR Generalizable Spoken Language Understanding, EMNLP 2024
YeonJoon Jung, Jaeseong Lee, Seungtaek Choi, Dohyeon Lee, Minsoo Kim, Seung-won Hwang. - Script-mix: Mixing Scripts for Low-resource Language Parsing, NAACL 2024
Jaeseong Lee, Dohyeon Lee, Seung-won Hwang. - ContrastiveMix: Overcoming Code-Mixing Dilemma in Cross-Lingual Transfer for Information Retrieval, NAACL 2024 (short)
Junggeun Do, Jaeseong Lee, Seung-won Hwang. - COMMIT: Code-Mixing English-Centric Large Language Model for Multilingual Instruction Tuning, Findings of NAACL 2024 (short)
Jaeseong Lee, YeonJoon Jung, Seung-won Hwang. - Multilingual Lottery Tickets to Pretrain Language Models, Findings of EMNLP 2023
Jaeseong Lee, Seung-won Hwang. - C2LIR: Continual Cross-lingual Transfer for Low-Resource Information Retrieval, ECIR 2023 (short)
{Jaeseong Lee, Dohyeon Lee}, Jongho Kim, Seung-won Hwang. - Script, Language, Labels: Overcoming Three Discrepancies for Low-Resource Language Specialization, AAAI 2023
Jaeseong Lee, Dohyeon Lee, Seung-won Hwang. - FAD-X: Fusing Adapters for Cross-lingual Transfer to Low-Resource Languages, AACL/IJCNLP 2022 (short)
Jaeseong Lee, Seung-won Hwang, Taesup Kim. - SCOPA: Soft Code-Switching, Pairwise Alignment for Zero-Shot Cross-lingual Transfer, CIKM 2021 (short)
{Dohyeon Lee, Jaeseong Lee}, Gyewon Lee, Byung-Gon Chun, Seung-won Hwang.