WHAT NEW IN X10.29 THU

AI collections from X · 22 collections · 58,925 posts

Back to No.001 GPT-6 Astra in Practice: Tests, Use Cases, and FeedbackUpdated 13 hours ago
gpt-6IMAGE

With ARC AGI 3 saturated, a proposal to use Millennium Prize problems as the new benchmark

The author jokes that GPT-6 Astra has already saturated ARC AGI 3, so the Clay Mathematics Institute's Millennium Prize problems should be the new benchmark, noting Astra and Anthropic's Fable both reach only about 4%.

samagra14
@samagra_sharmaArchived 2026.09.14Originally posted 2026.09.04
With ARC AGI 3 saturated, a proposal to use Millennium Prize problems as the new benchmark
samagra14 @samagra_sharmaOriginally posted 2026.09.04
Open original on X

Introducing ARC AGI 4 Now that GPT-6 Astra has already saturated the Arc AGI 3 benchmarks, it is time we introduce a newer more foundational benchmark to measure frontier intelligence. @AnthropicAI Fable and @ChatGPT Astra are both tied at ~4%. https://www.claymath.org/millennium-problems/

x.com/samagra_sharma/status/2095743467435761886Archived 2026.09.14 15:30
Open original on X