NEWS
 Vision-Language Models Failed the Core World Modeling TestThrough 660 experiments, it was revealed that fifteen contemporary vision-language models exhibited near-random accuracy in motion and spatial perception, indicating a significant gap from human-level world modeling. This is a news summary compiled for the AiPulse newsletter — not the original article. The original article is at aclanthology.org. Read the original article AI Critique → English (automatic translation) |