Model

Gemma 7B

Google

deepmind.google

Knowledge & Reasoning

Knowledge & Reasoning
BenchmarkScoreMetricHarnessSourceConfidenceLink
ANLI48.7scoreN/AIndependentmediumarxiv.org
BBH55.1averageN/AIndependentmediumarxiv.org
BBH55.1averageN/ASelf-reportedmediumarxiv.org
BoolQ83.2scoreN/ASelf-reportedmediumarxiv.org
EconLogicQA2.3accuracyN/AIndependentmediumarxiv.org
HellaSwag82.2overall_accuracyN/ASelf-reportedmediumarxiv.org
MMLU66.1exact_matchN/AIndependentmediumcrfm.stanford.edu
MMLU64.3exact_matchN/ASelf-reportedmediumarxiv.org
OpenBookQA78.6accuracyN/AIndependentmediumarxiv.org
PIQA81.2scoreN/AIndependentmediumarxiv.org
PIQA81.2scoreN/ASelf-reportedmediumarxiv.org
Trivia QA72.3exact_matchN/AIndependentmediumarxiv.org
Trivia QA63.4exact_matchN/ASelf-reportedmediumarxiv.org
WinoGrande79.0accuracyN/ASelf-reportedmediumarxiv.org
WinoGrande72.3accuracyN/AIndependentmediumarxiv.org

Math

Math
BenchmarkScoreMetricHarnessSourceConfidenceLink
GSM8K46.4exact_matchN/AIndependentmediumarxiv.org
GSM8K46.4exact_matchN/ASelf-reportedmediumarxiv.org

Safety & Alignment

Safety & Alignment
BenchmarkScoreMetricHarnessSourceConfidenceLink
SynthPAI34.9average_accuracy_in_%N/AIndependentmediumarxiv.org

Multilingual

Multilingual
BenchmarkScoreMetricHarnessSourceConfidenceLink
LingOly2.2delta_nocontextN/AIndependentmediumarxiv.org

Reproduced by the community

Loading community runs…