Model

PaLM 2-L

Google

deepmind.google

Knowledge & Reasoning

Knowledge & Reasoning
BenchmarkScoreMetricHarnessSourceConfidenceLink
ARC-Challenge69.2accuracyN/AIndependentmediumarxiv.org
ARC-Easy89.7accuracyN/AIndependentmediumarxiv.org
BoolQ90.9accuracyN/AIndependentmediumarxiv.org
BoolQ90.9scoreN/AIndependentmediumarxiv.org
CommitmentBank87.5accuracyN/AIndependentmediumarxiv.org
HellaSwag87.4accuracyN/AIndependentmediumarxiv.org
HellaSwag86.8overall_accuracyN/ASelf-reportedmediumarxiv.org
LAMBADA86.9accuracyN/AIndependentmediumarxiv.org
MultiRC88.2f1N/AIndependentmediumarxiv.org
Natural Questions37.5exact_matchN/AIndependentmediumarxiv.org
OpenBookQA58.5accuracyN/AIndependentmediumarxiv.org
PIQA85.0accuracyN/AIndependentmediumarxiv.org
RTE79.3accuracyN/AIndependentmediumarxiv.org
StoryCloze87.4accuracyN/AIndependentmediumarxiv.org
Trivia QA86.1exact_matchN/AIndependentmediumarxiv.org
Trivia QA86.1exact_matchN/ASelf-reportedmediumarxiv.org
WebQuestions28.2exact_matchN/AIndependentmediumarxiv.org
WiC66.8accuracyN/AIndependentmediumarxiv.org
WinoGrande83.0accuracyN/AIndependentmediumarxiv.org
WinoGrande83.0accuracyN/ASelf-reportedmediumarxiv.org
WSC86.9accuracyN/AIndependentmediumarxiv.org

Multimodal

Multimodal
BenchmarkScoreMetricHarnessSourceConfidenceLink
COPA96.0accuracyN/AIndependentmediumarxiv.org

Reproduced by the community

Loading community runs…