[KongchangAI]
Benchmark

SuperGPQA

A large-scale scientific knowledge question-answering benchmark used to evaluate AI models on reasoning and knowledge capabilities across multiple scientific domains.

Timeline (last 90 days)

Oct 2

On the SuperGPQA benchmark, the minimal baseline answered only 54.0% of questions correctly on the first attempt, while occupational profiles reached 71.6%, but this advantage stems from prompt length or format rather than domain expertise.

Unverified50%
Oct 2

SuperGPQA is a large-scale graduate-level knowledge question-answering benchmark covering 285 sub-disciplines, with questions primarily in the closed-ended single-choice format.

Unverified50%

All Facts (2)

Source Articles