WHAT NEW IN X11.03 TUE

AI collections from X · 22 collections · 66,427 posts

Back to No.001 GPT-6 Astra in Practice: Tests, Use Cases, and FeedbackUpdated 2 min ago
gpt-6THREAD

GPT-6 Astra cannot run without chain-of-thought, even at its lowest effort

The author found that Astra's chain-of-thought cannot be turned off and its lowest effort setting is 'low', yet it still reaches 99.6% multiplication accuracy.

Maksym Andriushchenko
@maksym_andrArchived 2026.09.17Originally posted 2026.09.17
THREADPosted in 2 parts

  1. Fun fact: GPT-6-Astra cannot be even run without CoT! The lowest possible effort is 'low' which leads to almost perfect multiplication accuracy, 99.6%!

    GPT-6 Astra cannot run without chain-of-thought, even at its lowest effort (thread part 1)

  2. Can frontier LLMs multiply numbers without using external tools? Absolutely. Claude Opus 5 gets 100% accuracy on multiplications up to 20x20 digits (1200/1200 correct). This accuracy does require max reasoning, however! It's an extremely simple eval, but it becomes relevant again for measuring no-CoT capabilities, particularly of recurrent-depth architectures (such as GPT-6-Astra).

    GPT-6 Astra cannot run without chain-of-thought, even at its lowest effort (thread part 2)
Maksym Andriushchenko @maksym_andrOriginally posted 2026.09.17
Open original on X
x.com/maksym_andr/status/2100275591811412019Archived 2026.09.17 10:17
Open original on X