Model

phi-1.5-web 1.3B

Microsoft · open weights

azure.microsoft.com

Knowledge & Reasoning

Knowledge & Reasoning
BenchmarkScoreMetricHarnessSourceConfidenceLink
ARC-Challenge44.9accuracyN/AIndependentmediumarxiv.org
ARC-Easy76.1accuracyN/AIndependentmediumarxiv.org
MMLU37.9average_(%)N/AIndependentmediumarxiv.org
siqa53.0accuracyN/AIndependentmediumarxiv.org
WinoGrande74.0accuracyN/AIndependentmediumarxiv.org

Coding

Coding
BenchmarkScoreMetricHarnessSourceConfidenceLink
MBPP43.5accuracyN/AIndependentmediumarxiv.org

Reproduced by the community

Loading community runs…