1
VLMs May Not Globally Enhance Human Alignment over LLMs During Natural Reading
VLM在自然阅读场景中未必比LLM更符合人类偏好，一项严谨的实验发现颠覆直观认知。
arXiv:2605.28818v1 Announce Type: new Abstract: Large language models (LLMs) have become increasingly useful computational models of human language pr…