Model

Mixtral-8x7B-Instruct-v0.1

Mistral AI · open weights

mistral.ai

Knowledge & Reasoning

Knowledge & Reasoning
BenchmarkScoreMetricHarnessSourceConfidenceLink
ANLI55.2scoreN/AIndependentmediumarxiv.org
GPQA Diamond30.6accuracyN/AIndependenthighepoch.ai
HellaSwag86.7overall_accuracyN/ASelf-reportedmediumarxiv.org
MMLU70.6exact_matchN/ASelf-reportedmediumarxiv.org
MMLU70.5exact_matchN/AIndependentmediumarxiv.org
OpenBookQA85.8accuracyN/AIndependentmediumarxiv.org
PIQA83.6scoreN/ASelf-reportedmediumarxiv.org
Trivia QA82.2exact_matchN/AIndependentmediumarxiv.org
WinoGrande77.2accuracyN/ASelf-reportedmediumarxiv.org
WinoGrande62.0accuracyN/AIndependentmediumarxiv.org

Math

Math
BenchmarkScoreMetricHarnessSourceConfidenceLink
GSM8K74.4exact_matchN/ASelf-reportedmediumarxiv.org

Reproduced by the community

Loading community runs…