arXivWeihan Peng, Chenxu Zhang, Qianao Wang, Yuling Shi, Heng Lian, Qihong Mao, Jiahao Pang, Chunliang Feng, Bowen Li, Xiaodong GuThu, May 28, 2026, 8:08 AM PDT
score 14.7
New test measures whether AI agents behave like consistent humans
Original: HEART-Bench: Do LLM Agents Exhibit Human-like Psychology?
Source: arxiv.org ↗
Writing ELI5 summary…