1 paper · 1 filter
Samidh Chakrabarti, David Willner, Kevin Klyman +3
This paper details the methodology behind CoPE, a policy-steerable small language model capable of fast and accurate content labeling. We present a novel training curricula called…