_horror
@__0xhorror__
We do not like the deception but respect the artistry
参加 November 2025
3.2K フォロー中 907 ファン
Yes that is much better. 80% lol
Just completed a full end to end run of the full 113 task DeepSWE benchmark on ox-alpha. The Rumored ~80% pass rate is completely incorrect. Actual benchmark result is 58.4%, landing the model at almost identical performance to Claude Opus 4.8 (59%) 🧵
もっと見る
