STILL: Selecting Tokens for Intra-Layer Hybrid Attention to Linearize LLMs

Add code
Feb 02, 2026

Share this with someone who'll enjoy it:

View paper onarxiv icon

Share this with someone who'll enjoy it: