Prompt Compression in the Wild: Measuring Latency, Rate Adherence, and Quality for Faster LLM Inference

Fuente: arXiv

Saved in:

Bibliographic Details
Main Authors:	Kummer, Cornelius, Jurkschat, Lena, Färber, Michael, Vahdati, Sahar
Format:	Preprint
Published:	2026
Subjects:	Information Retrieval Artificial Intelligence Computation and Language
Online Access:	Acceder al recurso
Tags:	Add Tag No Tags, Be the first to tag this record!

Internet

https://arxiv.org/abs/2604.02985