gpt-6POST
GPT-6 Astra overturns expert answers on physics benchmark, score rises to 98%
On a physics benchmark, GPT-6 Astra insisted the expert answers were wrong and its own calculations were right; a team checked and confirmed Astra was correct, raising its score from 80% to 98%.
Freddy Vega
@freddierArchived 16 days agoOriginally posted 2026.10.15
Freddy Vega @freddierOriginally posted 2026.10.15
GPT-6 Astra got +80% on a Physics benchmark. It insisted the expert results were wrong and its calculations were right. A team checked. And yes. GPT-6 Astra was right. New score: 98%
This post quotes another post: x.com/pratikrath15/status/2110225213401206892
x.com/freddier/status/2110443503516057682Archived 2026.10.15 10:51