DAM: Dynamic Attention Mask for Long-Context Large Language Model Inference Acceleration.

Hanzhi Zhang, Heng Fan, Kewei Sha, Yan Huang, Yunhe Feng

Browse the full ACL paper archive.