Reiichiro Nakano

AI

The reported MATH-500 score rose from 7.9% to 48.1% with the same sampling settings.

18240123456789012345678901234567890123456789

TW

AK

RO

226 Sources, 4d ago, updated 4d ago

AI

In the announcement, @bneyshabur places Mirendil-Base in the 100B–500B-parameter range and promises independent evaluations.

466012345678901234567890123456789

CM

TR

BN

74 Sources, 14d ago, updated 14d ago

AI

A researcher reports GPT-6 Astra completing a Minecraft task via computer use.

8801234567890123456789

DE

RN

WZ

4 Sources, 65d ago, first seen 65d ago

AI

Replies post benchmark tables and charts comparing the models on internal tasks.

8201234567890123456789

RS

DF

PW

49 Sources, 67d ago, first seen 67d ago

AI

Tweet describes 404 errors and downtime tracker showing broad chatbot issues.

3701234567890123456789

MB

DF

JE

30 Sources, 67d ago, first seen 67d ago