1 paper
Rotem Dror, Zohar Elyoseph, Yuval Haber +3
Existing approaches to persona simulation with Large Language Models (LLMs) mostly rely on shallow character descriptions that fail to sustain coherent character behavior across ex…