Do large language models and humans have similar behaviors in causal inference with script knowledge?

Re-assign community