SALE : Low-bit Estimation for Efficient Sparse Attention in Long-context LLM Prefilling

Add code
May 30, 2025

Share this with someone who'll enjoy it:

View paper onarxiv icon

Share this with someone who'll enjoy it: