Model

LLaMA 65B

Meta · open weights

ai.meta.com

Knowledge & Reasoning

Knowledge & Reasoning
BenchmarkScoreMetricHarnessSourceConfidenceLink
ARC-Challenge56.0accuracyN/AIndependentmediumarxiv.org
ARC-Easy78.9accuracyN/AIndependentmediumarxiv.org
BBH58.4averageN/AIndependentmediumarxiv.org
BBH43.5averageN/ASelf-reportedmediumarxiv.org
BoolQ87.1scoreN/AIndependentmediumcrfm.stanford.edu
BoolQ85.3accuracyN/AIndependentmediumarxiv.org
BoolQ85.3scoreN/ASelf-reportedmediumarxiv.org
HellaSwag84.2accuracyN/AIndependentmediumarxiv.org
HellaSwag84.2overall_accuracyN/ASelf-reportedmediumarxiv.org
LAMBADA77.7scoreN/AIndependentmediumarxiv.org
MMLU63.4average_(%)N/AIndependentmediumarxiv.org
MMLU63.4exact_matchN/ASelf-reportedmediumarxiv.org
Natural Questions39.9exact_matchN/AIndependentmediumarxiv.org
Natural Questions35.0exact_matchN/AIndependentmediumarxiv.org
Natural Questions31.0exact_matchN/AIndependentmediumarxiv.org
OpenBookQA60.2accuracyN/ASelf-reportedmediumarxiv.org
OpenBookQA60.2accuracyN/AIndependentmediumarxiv.org
PIQA82.8accuracyN/AIndependentmediumarxiv.org
PIQA82.8scoreN/ASelf-reportedmediumarxiv.org
PIQA82.8scoreN/AIndependentmediumarxiv.org
siqa52.3accuracyN/AIndependentmediumarxiv.org
Trivia QA86.0exact_matchN/ASelf-reportedmediumarxiv.org
Trivia QA73.0exact_matchN/AIndependentmediumarxiv.org
Trivia QA72.6exact_matchN/AIndependentmediumarxiv.org
Trivia QA71.6exact_matchN/AIndependentmediumarxiv.org
Trivia QA68.2exact_matchN/AIndependentmediumarxiv.org
WinoGrande77.0accuracyN/AIndependentmediumarxiv.org
WinoGrande77.0accuracyN/ASelf-reportedmediumarxiv.org

Math

Math
BenchmarkScoreMetricHarnessSourceConfidenceLink
GSM8K54.4exact_matchN/AIndependentmediumarxiv.org
GSM8K50.9accuracyN/AIndependentmediumarxiv.org
GSM8K50.9exact_matchN/ASelf-reportedmediumarxiv.org
MATH10.6accuracyN/AIndependentmediumarxiv.org

Coding

Coding
BenchmarkScoreMetricHarnessSourceConfidenceLink
MBPP37.7accuracyN/AIndependentmediumarxiv.org

Reproduced by the community

Loading community runs…