Sewon Min

sewon

·

https://shmsw25.github.io/

AI & ML interests

natural language processing, language modeling

Recent Activity

upvoted a paper 1 day ago

Scaling Repeated-Data Pretraining: Batch Size, Regularization, and GPU-Time Tradeoffs

submitted a paper 1 day ago

Scaling Repeated-Data Pretraining: Batch Size, Regularization, and GPU-Time Tradeoffs

upvoted a paper 5 months ago

EMO: Pretraining Mixture of Experts for Emergent Modularity

View all activity

Organizations

Papers 28

arxiv:2511.19399

arxiv:2507.01297

arxiv:2506.10947

arxiv:2506.08276

models 0

None public yet

datasets 1

sewon/ambig_qa

Viewer • Updated Jan 9, 2024 • 24.1k • 1.47k • 21