Model

PaLM 540B

Google

deepmind.google

Knowledge & Reasoning

Knowledge & Reasoning
BenchmarkScoreMetricHarnessSourceConfidenceLink
BoolQ88.7scoreN/AIndependentmediumarxiv.org
HellaSwag83.8accuracyN/AIndependentmediumarxiv.org
HellaSwag83.8overall_accuracyN/ASelf-reportedmediumjmlr.org
HellaSwag83.6accuracyN/AIndependentmediumarxiv.org
HellaSwag83.4accuracyN/AIndependentmediumarxiv.org
LAMBADA89.7accuracyN/AIndependentmediumarxiv.org
LAMBADA81.8accuracyN/AIndependentmediumarxiv.org
LAMBADA77.9accuracyN/AIndependentmediumarxiv.org
LAMBADA77.9scoreN/AIndependentmediumarxiv.org
MMLU69.3exact_matchN/AIndependentmediumarxiv.org
Natural Questions39.6exact_matchN/AIndependentmediumarxiv.org
Natural Questions29.3exact_matchN/AIndependentmediumarxiv.org
Natural Questions21.2exact_matchN/AIndependentmediumarxiv.org
OpenBookQA68.0accuracyN/ASelf-reportedmediumjmlr.org
OpenBookQA53.4accuracyN/AIndependentmediumarxiv.org
OpenBookQA53.4accuracyN/AIndependentmediumarxiv.org
PIQA82.3scoreN/AIndependentmediumarxiv.org
RTE79.6accuracyN/AIndependentmediumarxiv.org
RTE78.7accuracyN/AIndependentmediumarxiv.org
RTE72.9accuracyN/AIndependentmediumarxiv.org
StrategyQA76.4accuracyN/AIndependentmediumarxiv.org
Trivia QA81.4exact_matchN/AIndependentmediumarxiv.org
Trivia QA81.4exact_matchN/AIndependentmediumarxiv.org
Trivia QA81.4exact_matchN/ASelf-reportedmediumjmlr.org
Trivia QA76.9exact_matchN/AIndependentmediumarxiv.org
WebQuestions43.5exact_matchN/AIndependentmediumarxiv.org
WebQuestions22.6exact_matchN/AIndependentmediumarxiv.org
WebQuestions10.6exact_matchN/AIndependentmediumarxiv.org
WinoGrande85.1accuracyN/AIndependentmediumarxiv.org
WinoGrande85.1accuracyN/ASelf-reportedmediumjmlr.org
WinoGrande81.1accuracyN/AIndependentmediumarxiv.org
WSC89.5accuracyN/AIndependentmediumarxiv.org
WSC89.1accuracyN/AIndependentmediumarxiv.org
WSC86.3accuracyN/AIndependentmediumarxiv.org

Math

Math
BenchmarkScoreMetricHarnessSourceConfidenceLink
GSM8K56.5accuracyN/AIndependentmediumarxiv.org
GSM8K56.5exact_matchN/AIndependentmediumarxiv.org
GSM8K17.9accuracyN/AIndependentmediumarxiv.org
MATH8.8accuracyN/AIndependentmediumarxiv.org

Coding

Coding
BenchmarkScoreMetricHarnessSourceConfidenceLink
MBPP36.8accuracyN/AIndependentmediumarxiv.org

Multilingual

Multilingual
BenchmarkScoreMetricHarnessSourceConfidenceLink
MGSM55.0average_(%)N/AIndependentmediumarxiv.org
TyDiQA-GoldP52.9exact_matchN/AIndependentmediumarxiv.org

Reproduced by the community

Loading community runs…