AI RESEARCH
Activation Compression in LLMs: Theoretical Analysis and Efficient Algorithm
arXiv CS.LG
•
ArXi:2605.01255v1 Announce Type: new