1 paper · 1 filter
Patryk Bartkowiak, Michal Podstawski
This paper introduces a novel dataset and evaluation benchmark designed to assess and improve small language models deployable on edge devices, with a focus on user profiling from…