इसका टेक्स्ट मैसेज भेजे: Can large language model agents simulate human trust behaviors?