Mind the Gap: The Divergence Between Human and LLM-Generated Tasks

Yi-Long Lu, Jiajun Song, Chunhui Zhang +1

Humans constantly generate a diverse range of tasks guided by internal motivations. While generative agents powered by large language models (LLMs) aim to simulate this complex beh…