Share your thoughts, 1 month free Claude Pro on usSee more
WorkDL logo mark

GEM

Benchmarks

Task NameDataset NameSOTA ResultTrend
Grounded ECG InterpretationGEM (test)
Diagnosis Accuracy83.3
18
Logical ReasoningGEM mastermind v1 (test)
Pass@178.6
6
Logical reasoning and game playingGEM
Sudoku pass@199.3
6
RDF-to-text generationGEM2 Factual (test)
Grammaticality98.4
6
RDF-to-text generationGEM2 Fictional (test)
Grammaticality93.7
6
Showing 5 of 5 rows