Teortaxes Notes Eval's Limited Lifespan
The commentator states it will still distinguish weaker models temporarily.
TLDR
Teortaxes, identified as a pseudonymous AI developer and commentator primarily recognized as DeepSeek's most vocal online booster, made the observation that an evaluation had been great while it lasted. According to the post, this benchmark will continue serving to discriminate against lesser models for a while longer. The statement comes from public activity associated with the creator account. No further details on the specific evaluation or reasons for its change in status appear in the available source material.
Combined views
8.3K
2 Sources, first seen 27d ago