Văn bản này: Can large language model agents simulate human trust behaviors?