14 lines
300 B
Markdown
14 lines
300 B
Markdown
---
|
|
title: "Song Han"
|
|
created: 2026-05-14
|
|
updated: 2026-05-14
|
|
type: entity
|
|
tags: [researcher, author]
|
|
---
|
|
|
|
# Song Han
|
|
|
|
- **机构**: MIT / NVIDIA
|
|
- **研究方向**: Efficient deep learning (TinyML, Deep Compression), model acceleration
|
|
- **代表工作**: [[streaming-llm|StreamingLLM]] (ICLR 2024)
|