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

D2

Benchmarks

Task NameDataset NameSOTA ResultTrend
ClassificationD2
Mean Accuracy91.611
30
Aspect Sentiment Triplet ExtractionD2 (16Res)
F1 Score74.83
25
Aspect Sentiment Triplet ExtractionD2 15Res
F1 Score66.12
25
Aspect Sentiment Triplet ExtractionD2 14Lap
F1 Score63.61
25
Aspect Sentiment Triplet ExtractionD2 14Res
F1 Score75.59
25
Spatio-temporal signal recoveryD2
MAE0.12
18
Time Series ForecastingD2 Synthetic (test)
MSE0.599
16
Medical Image SegmentationD2
DSC87.35
14
PCB RoutingD2 Synthetic Gridless Board
CP100
10
RegressionD2
RMSE0.16
10
RegressionD2
Average Relative MSE0.084
10
ClassificationD2 0.15 (test)
Mean Accuracy91.657
10
ICD-10 Code PredictionD2 noisy (test)
AUPRC (Z37)93.86
10
HDI Tensor CompletionD2
MAE3.5514
9
Outlier DetectionD2 with only clusteriers (test)
AUC0.918
9
Aspect-level sentiment classificationD2
Accuracy72.08
9
Knee cartilage segmentationD2
Dice94.14
7
Root Cause LocalizationD2 complete data conditions
Top-1 Accuracy81.5
7
Failure TriageD2 complete data conditions
Precision88.2
6
Anomaly DetectionD2 complete data conditions
Precision99.3
6
Time-Domain PredictionD2
NMSE (dB)-18.58
6
ProcessingD2 Operational JSON
Processing Time (s)18.03
5
ProcessingD2 (Operational) CSV
Processing Time (s)12.55
5
Expert-grounded reasoning alignmentD2 N=499
GT-HYB81.3
5
Reliability AssessmentD2 (test)
AU-ARC92.1
5
Showing 25 of 37 rows