Aditya Somasundaram

~Aditya_Somasundaram1

1

论文总数

1.0

年均投稿

7.3

平均评分

接收情况1/1

会议分布

NeurIPS

1

发表论文 (1 篇)

20251 篇

Small Batch Size Training for Language Models: When Vanilla SGD Works, and Why Gradient Accumulation is Wasteful

NeurIPS 2025Poster

合作者 (4)

Andrew Gordon Wilson1 篇

Martin Marek1 篇

Micah Goldblum1 篇

Sanae Lotfi1 篇