List your company

Classification on HotpotQA RAG v3

90.75Macro F1SURE-RAG calibrated
15.589235.102154.61574.1279May 5, 2026

Evaluation results

MethodLinks
SURE-RAG calibratedn=32026.0590.7586.6295.6989.94
SURE-RAGn=32026.0589.5184.6595.0888.78
Concat cross-encodern=32026.0588.8884.9694.3787.31
SURE pooling-onlyn=32026.0583.8377.8893.9379.67
DeBERTa mean-pooln=32026.0565.1654.4473.0767.98
DeBERTa top-k pooln=32026.0564.4454.1771.3667.78
Overlap-onlyn=12026.0561.0163.9467.7551.33
DeBERTa max-pooln=32026.0559.4832.0379.2767.15
Concat TF-IDFn=12026.0535.0115.5830.9558.48
Majorityn=12026.0518.480055.45