Shiwei Liu

Shiweiliuiiiiiii

·

https://shiweiliuiiiiiii.github.io/

Shiwei_Liu66

AI & ML interests

LLM, pre-training, post-training, RL, efficient AI

Recent Activity

upvoted a paper 1 day ago

Looping Beyond Twice: A Scalable Recipe for Looped Mixture-of-Experts

submitted a paper 1 day ago

Looping Beyond Twice: A Scalable Recipe for Looped Mixture-of-Experts

upvoted a paper 5 days ago

FlashLoop: Fast and Memory-Efficient Looped Transformers via Lazy Updates

View all activity

Organizations

None yet

Papers 18

arxiv:2502.20545

arxiv:2502.17055

arxiv:2502.07490

arxiv:2502.05795

models 5

Shiweiliuiiiiiii/Llama-3.1-8B-EdgeLLM

Updated Nov 20, 2024

Shiweiliuiiiiiii/mixln-1b-6posts

Updated Oct 27, 2024

Shiweiliuiiiiiii/250m-postLN

Updated Oct 25, 2024 • 6

Shiweiliuiiiiiii/250m-preLN

Updated Oct 25, 2024 • 3

Shiweiliuiiiiiii/mixln_1b_6posts

Updated Oct 20, 2024 • 9

datasets 0

None public yet