WHAT NEW IN X11.07 SAT

AI collections from X · 23 collections · 70,791 posts

Back to No.001 GPT-6 Astra in Practice: Tests, Use Cases, and FeedbackUpdated 14 hours ago
gpt-6IMAGE

ValsAI tests Astra and Opus at every reasoning level on math proofs

ValsAI ran GPT-6 Astra, Opus 5 and Opus 5.5 at every reasoning level, from low to max, on its Proof Bench v1.1 benchmark.

Vals AI
@ValsAIArchived 2026.09.25Originally posted 2026.09.25
ValsAI tests Astra and Opus at every reasoning level on math proofs
Vals AI @ValsAIOriginally posted 2026.09.25
Open original on X

How does "thinking harder" change how models perform on math proofs? We ran GPT-6 Astra, Opus 5, and Opus 5.5 on every reasoning level: low, medium, high, xhigh, and max on our benchmark, Proof Bench v1.1.

x.com/valsai/status/2103283609704837290Archived 2026.09.25 10:09
Open original on X