Model

PaLM 2-S

Google

deepmind.google

Knowledge & Reasoning

Knowledge & Reasoning
BenchmarkScoreMetricHarnessSourceConfidenceLink
ARC-Challenge59.6accuracyN/AIndependentmediumarxiv.org
ARC-Easy85.6accuracyN/AIndependentmediumarxiv.org
BoolQ88.1accuracyN/AIndependentmediumarxiv.org
BoolQ88.1scoreN/AIndependentmediumarxiv.org
CommitmentBank82.1accuracyN/AIndependentmediumarxiv.org
HellaSwag85.6accuracyN/AIndependentmediumarxiv.org
HellaSwag82.0overall_accuracyN/ASelf-reportedmediumarxiv.org
LAMBADA80.7accuracyN/AIndependentmediumarxiv.org
MultiRC84.0f1N/AIndependentmediumarxiv.org
Natural Questions25.3exact_matchN/AIndependentmediumarxiv.org
OpenBookQA57.4accuracyN/AIndependentmediumarxiv.org
OpenBookQA56.2accuracyN/ASelf-reportedmediumarxiv.org
PIQA82.2accuracyN/AIndependentmediumarxiv.org
RTE78.7accuracyN/AIndependentmediumarxiv.org
StoryCloze85.6accuracyN/AIndependentmediumarxiv.org
Trivia QA75.2exact_matchN/AIndependentmediumarxiv.org
Trivia QA75.2exact_matchN/ASelf-reportedmediumarxiv.org
WebQuestions21.8exact_matchN/AIndependentmediumarxiv.org
WiC50.6accuracyN/AIndependentmediumarxiv.org
WinoGrande77.9accuracyN/AIndependentmediumarxiv.org
WinoGrande77.9accuracyN/ASelf-reportedmediumarxiv.org
WSC84.6accuracyN/AIndependentmediumarxiv.org

Coding

Coding
BenchmarkScoreMetricHarnessSourceConfidenceLink
MBPP50.0accuracyN/AIndependentmediumarxiv.org

Multimodal

Multimodal
BenchmarkScoreMetricHarnessSourceConfidenceLink
COPA89.0accuracyN/AIndependentmediumarxiv.org

Instruction Following

Instruction Following
BenchmarkScoreMetricHarnessSourceConfidenceLink
IFEval59.1inst-level_loose-accuracyN/AIndependentmediumarxiv.org

Reproduced by the community

Loading community runs…