Gavia Gray

~Gavia_Gray1

3

论文总数

1.5

年均投稿

6.2

平均评分

接收情况3/3

会议分布

NeurIPS

2

ICLR

1

发表论文 (3 篇)

20252 篇

Power Lines: Scaling laws for weight decay and batch size in LLM pre-training

NeurIPS 2025Poster

Straight to Zero: Why Linearly Decaying the Learning Rate to Zero Works Best for LLMs

ICLR 2025Poster

20241 篇

Normalization Layer Per-Example Gradients are Sufficient to Predict Gradient Noise Scale in Transformers

NeurIPS 2024Poster

合作者 (6)

Joel Hestness3 篇

Shane Bergsma3 篇

Daria Soboleva2 篇

Gurpreet Gosal2 篇

Nolan Simran Dey2 篇

Aman Tiwari1 篇