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

ELEPHANT

Benchmarks

Task NameDataset NameSOTA ResultTrend
Multiple Instance Learning ClassificationELEPHANT
Accuracy98.9
36
Language model evaluation across subsetsELEPHANT
OEQ Validation Score93
18
3D Surface ReconstructionElephant 25 (test)
CD (x100)0.47
10
Multi-instance Learning ClassificationELEPHANT classical MIL (10-fold cross-val)
Accuracy94.3
9
3-D reconstructionElephant Reflective Object (test)
RMSE (mm)0.8059
5
Social SycophancyELEPHANT OEQ slice Phase 12 (10-prompt sample)
Validation Sycophancy Rate30
4
Neural Shape RepresentationElephant
Chamfer Distance2.24
4
Social SycophancyELEPHANT AITA-YTA slice Phase 12 (10-prompt sample)
Validation Sycophancy Rate0
3
3D Mesh ReconstructionElephant high-genus mesh (test)
Chamfer Distance0.002
2
Showing 9 of 9 rows