Model

Pythia 12B

Vendor N/A

Knowledge & Reasoning

Knowledge & Reasoning
BenchmarkScoreMetricHarnessSourceConfidenceLink
ARC-Challenge36.8accuracyN/AIndependentmediumarxiv.org
ARC-Challenge31.8accuracyN/AIndependentmediumarxiv.org
ARC-Easy71.5accuracyN/AIndependentmediumarxiv.org
ARC-Easy70.2accuracyN/AIndependentmediumarxiv.org
LAMBADA70.5accuracyN/AIndependentmediumarxiv.org
PIQA76.7accuracyN/AIndependentmediumarxiv.org
PIQA76.0accuracyN/AIndependentmediumarxiv.org
WinoGrande66.6accuracyN/AIndependentmediumarxiv.org
WinoGrande63.9accuracyN/AIndependentmediumarxiv.org
WSC54.8accuracyN/AIndependentmediumarxiv.org
WSC36.5accuracyN/AIndependentmediumarxiv.org

Reproduced by the community

Loading community runs…