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

HRS

Benchmarks

Task NameDataset NameSOTA ResultTrend
Expressive EvaluationHRS benchmark
Creativity66.97
21
10-year MortalityHRS (held-out set)
AUC0.778
16
Layout predictionHRS spatial
Accuracy86.07
11
Layout predictionHRS numerical
Precision93.28
11
Text-to-Image GenerationHRS
Count F166
10
Spatial ReasoningHRS
Accuracy53.96
8
Numerical ReasoningHRS
Precision78.65
8
Grounding AccuracyHRS
Spatial Accuracy45.01
8
GroundingHRS-Spatial
mIoU0.372
8
Prompt FidelityHRS dataset
CLIP Score33.63
6
Text-to-Image GenerationHRS benchmark
CLIP Score33.63
2
Showing 11 of 11 rows