MadMix-v0.2: benchmark results

Provider: Other.

Aggregate Rating 1411 ± 33, rank #5571 of 10916 rated models, from 6 benchmark results.

Strongest benchmark results

BenchmarkScoreMetricPercentile
Open LLM Leaderboard v1 - ARC Challenge64.85Normalized accuracy (%) (25-shot)68.1
Open LLM Leaderboard v1 - TruthfulQA MC255.79MC2 (%) (0-shot)66.2
Open LLM Leaderboard v1 - MMLU64.02Accuracy (%) (5-shot)64.9
Open LLM Leaderboard v1 - HellaSwag83.54Normalized accuracy (%) (10-shot)59.5
Open LLM Leaderboard v1 - WinoGrande77.35Accuracy (%) (5-shot)55.2
Open LLM Leaderboard v1 - GSM8K30.78Accuracy (%) (5-shot)47.7

Interactive version: theaggregate.ai/model?slug=madmix-v0-2 · How It Works · Data refreshed daily, snapshot 2026-11-08.