FlashLoop: Fast and Memory-Efficient Looped Transformers via Lazy Updates
Paper • 2609.29812 • Published • 3 •
1
Wanqi Yang
WanqiYang
·
AI & ML interests
None yet
Recent Activity
upvoted a paper 30 minutes ago
Looping Beyond Twice: A Scalable Recipe for Looped Mixture-of-Experts upvoted a paper 6 days ago
DepthBench: Measuring How Residual Connections Enable More Computational Depth commentedon a paper 10 days ago
FlashLoop: Fast and Memory-Efficient Looped Transformers via Lazy UpdatesOrganizations
commented a
paper 10 days ago