However, the new Anthropic model remains inaccessible to users.
However, the new Anthropic model remains inaccessible to users.
@EpochAIResearch 1 point? That's it? ALL THIS MADNESS, FEAR, APOCALYPTIC WARNINGS, IPO, NATIONALIZATION OVER 1 POINT? Ladies and gentlemen, this is the straw that broke the camel's back.
@GaryMarcus you have no clue what you are talking about, as usual.
@scaling01 Claude is by far the best in agentic coding
If you run Opus "as a model Fable hands off to", it does a 2x better job. (Is it 2x more tokens or just Fable pricing?) Such zany stuff always reminds me of Duck Tales episode where some villains hypnotized sturgeons into believing they're chicken, to harvest caviar daily https://x.com/teortaxesTex/status/2066742951628189723/photo/1 https://twitter.com/xeophon/status/2066741907644874973
@scaling01 I think it's a fair rating GPT 5.5 has very strong post-training and we see on MathArena that Fable isn't uniformly the strongest model even against normal 5.5. Pro can genuinely lead in some dimensions https://x.com/teortaxesTex/status/2066685034778562850/photo/1
Fable is slightly higher rated than GPT-5.5-Pro on Epoch's ECI I suspect as we get more benchmark results its ECI should improve to ~163 https://twitter.com/EpochAIResearch/status/2066674892809101767
don't worry guys, Mythos is not that good at coding insane score on ProgramBench https://x.com/scaling01/status/2066687980538904648/photo/1 https://twitter.com/scaling01/status/2066683927926853635
even when you truncate the x axis at 150, exaggerating the gains, this looks like gradual change rather than a whole new order. https://twitter.com/epochairesearch/status/2066674892809101767
trump and anthropic freaking out together, over what turns out to be a 1 point benchmark gain 🤦♂️ https://twitter.com/garymarcus/status/2066689308640055764
Ask a question below.
Published answers will appear here.
