BinghengWu's picture

3 6 10

BinghengWu

wubingheng

·

https://github.com/wubingheng111

AI & ML interests

I like to fine-tune the small models of the Doge series.

Organizations

authored 2 papers 3 months ago

Trainable Dynamic Mask Sparse Attention

Paper • 2508.02124 • Published Aug 4 • 17

Concise Reasoning, Big Gains: Pruning Long Reasoning Trace with Difficulty-Aware Prompting

Paper • 2505.19716 • Published May 26 • 4

authored 2 papers 10 months ago

Cheems: Wonderful Matrices More Efficient and More Effective Architecture

Paper • 2407.16958 • Published Jul 24, 2024 • 4

Wonderful Matrices: Combining for a More Efficient and Effective Foundation Model Architecture

Paper • 2412.11834 • Published Dec 16, 2024 • 8