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

GRAFITE

Benchmarks

Task NameDataset NameSOTA ResultTrend
Underspecified PromptingGRAFITE Sample
Pass Rate16.7
4
Table ProcessingGRAFITE Sample Dataset Table
Pass Rate54.5
4
SummarizationGRAFITE Sample
Pass Rate83.3
4
ReasoningGRAFITE Sample Reasoning
Pass Rate83.3
4
MultilingualGRAFITE Multilingual Sample Dataset
Pass Rate40
4
MathematicsGRAFITE Math Sample
Pass Rate60
4
Instruction FollowingGRAFITE Sample Instruction Following
Pass Rate58.3
4
Factual AnalysisGRAFITE Sample Factual
Pass Rate29.4
4
Creative WritingGRAFITE Sample Dataset Creative
Pass Rate100
4
CodeGRAFITE Sample Code
Pass Rate100
4
Multi-domain evaluationGRAFITE Sample Dataset (Total)
Pass Rate63.2
4
Showing 11 of 11 rows