Model

Nemotron-4 15B

NVIDIA · open weights

nvidia.com

Knowledge & Reasoning

Knowledge & Reasoning
BenchmarkScoreMetricHarnessSourceConfidenceLink
BBH58.7averageN/ASelf-reportedmediumarxiv.org
HellaSwag82.4overall_accuracyN/ASelf-reportedmediumarxiv.org
MMLU58.7exact_matchN/ASelf-reportedmediumarxiv.org
PIQA82.4scoreN/ASelf-reportedmediumarxiv.org
WinoGrande78.0accuracyN/ASelf-reportedmediumarxiv.org

Math

Math
BenchmarkScoreMetricHarnessSourceConfidenceLink
GSM8K46.0exact_matchN/ASelf-reportedmediumarxiv.org

Reproduced by the community

Loading community runs…