Morning Edition · Sunday, October 11, 2026Published at 2:16 AM EDT · New York

Scale's Humanity's Sixth Sense Benchmark: Best Model Scores 53.6% on Intuitive Visual Reasoning Against 93.1% for Humans

Of 8,573 model failures, 94% came from perception or inference errors and only 5% from faulty logic, which points to the visual encoder, not the reasoning stack, as the bottleneck.

Scale Labs and Elorian released Humanity's Sixth Sense (HSS) on October 7. It is an open-ended benchmark of 522 tasks built on 288 images and 234 video clips, with 17.6 hours of video in total. The tasks ask questions that people answer qui…