Convergence and Implicit Bias of Gradient Descent on Continual Linear Classification
ð¥ Hyunji Jung* , Hanseul Cho* , and Chulhee Yun
ð Under Review ð
ð Best Paper Award at JKAIA 2024
Publications
You can also find my articles on my Google Scholar profile.
Arithmetic Transformers Can Length-Generalize in Both Operand Length and Count
ð¥ Hanseul Cho* , Jaeyoung Cha* , Srinadh Bhojanapalli , and Chulhee Yun
ð Arxiv Preprint ð ð [paper]
Position Coupling: Improving Length Generalization of Arithmetic Transformers Using Task Structure
ð¥ Hanseul Cho* , Jaeyoung Cha* , Pranjal Awasthi , Srinadh Bhojanapalli , Anupam Gupta , and Chulhee Yun
ð NeurIPS 2024 (Short version at ICML 2024 Workshop on Long-Context Foundation Models (LCFM)) ð ð [paper]
DASH: Warm-Starting Neural Network Training in Stationary Settings without Loss of Plasticity
ð¥ Baekrok Shin* , Junsoo Oh* , Hanseul Cho , and Chulhee Yun
ð NeurIPS 2024 (Short version at ICML 2024 Workshop on Workshop on Advancing Neural Network Training (WANT)) ð ð [paper]
Fundamental Benefit of Alternating Updates in Minimax Optimization
ð¥ Jaewook Lee* , Hanseul Cho* , and Chulhee Yun
ð ICML 2024 (Short version at ICLR 2024 Workshop on Bridging the Gap Between Practice and Theory in Deep Learning (BGPT)) ð ð [paper]
ð Spotlight @ ICML 2024 (Top 3.5% among total submissions)
Fair Streaming Principal Component Analysis: Statistical and Algorithmic Viewpoint
ð¥ Junghyun Lee* , Hanseul Cho* , Se-Young Yun , and Chulhee Yun
ð NeurIPS 2023 ð ð [paper]
PLASTIC: Improving Input and Label Plasticity for Sample Efficient Reinforcement Learning
ð¥ Hojoon Lee* , Hanseul Cho* , Hyunseung Kim* , Daehoon Gwak , Joonkee Kim , Jaegul Choo , Se-Young Yun , and Chulhee Yun
ð NeurIPS 2023 ð ð [paper]
SGDA with shuffling: faster convergence for nonconvex-PÅ minimax optimization
ð¥ Hanseul Cho and Chulhee Yun
ð ICLR 2023 ð ð [paper]
ð NAVER Outstanding Theory Paper Award at 2022 KAIA-NAVER Autumnal Joint Conference