1 paper · 1 filter
Gurvan Richardeau, Erwan Le Merrer, Camilla Penzo +1
In a parallel with the 20 questions game, we present a method to determine whether two large language models (LLMs), placed in a black-box context, are the same or not. The goal is…