MadMix-v0.1: benchmark results

Provider: Other.

Aggregate Rating 1416 ± 33, rank #5339 of 10916 rated models, from 6 benchmark results.

Strongest benchmark results

BenchmarkScoreMetricPercentile
Open LLM Leaderboard v1 - MMLU64.37Accuracy (%) (5-shot)69
Open LLM Leaderboard v1 - ARC Challenge64.93Normalized accuracy (%) (25-shot)68.5
Open LLM Leaderboard v1 - HellaSwag84.37Normalized accuracy (%) (10-shot)65.7
Open LLM Leaderboard v1 - GSM8K49.66Accuracy (%) (5-shot)62.2
Open LLM Leaderboard v1 - WinoGrande77.19Accuracy (%) (5-shot)53.9
Open LLM Leaderboard v1 - TruthfulQA MC251.05MC2 (%) (0-shot)50.5

Interactive version: theaggregate.ai/model?slug=madmix-v0-1 · How It Works · Data refreshed daily, snapshot 2026-11-08.