Harrison Kinsley

@Sentdex · 7 nov. 2024

Grok is strong. Here are some more benchmark results. Continuing along from the Bigcodebench I did in the QT, here are the benchmarks on the HF Leaderboard (BBH, GPQA, IFEVAL, MathLevel5, MUSR, MMLUPro) + TrivQA, first with an overall, then heatmap per model and bench. Don't

Harrison Kinsley@Sentdex· 5 nov. 2024

Wow, just finished benching XAI's Grok-beta on a slightly modified internal Bigcodebench. This is much stronger than I was expecting.

Elon Musk

@elonmusk

Cool

04:31 · 7 novembre 2024 · 32 k vues

530

X →