X · @teortaxesTex
· X / Twitter
…K3 scores almost 300 Elo points above Fable on creative writing. But note: this is done with Claude Sonnet 4.6 as judge. Claude may be largely measu…
…K3 scores almost 300 Elo points above Fable on creative writing.But note: this is done with Claude Sonnet 4.6 as judge. Claude may be largely measuring Claude-ness. It so happens that K3 is more Claude than Claude itself has ever been.Sam Paech: Creative writing bench updates. We got:- kimi-k3- gpt-5.6- muse-spark-1.1