Scale Labs introduces Humanity's Sixth Sense visual reasoning benchmark

Scale Labs has introduced Humanity's Sixth Sense (HSS), a benchmark designed to evaluate intuitive visual reasoning. The suite spans 522 open-ended tasks across images and video, covering domains ranging from spatial and causal reasoning to social understanding.

According to Scale Labs, human evaluators achieved a 93.1% benchmark score. Among tested AI models, GPT-6-astra achieved the highest result at 53.6%, while the median model scored 30.9%.