Incorrectness detection on NQ-Open benchmark leaderboard · SOTA2 Research

Incorrectness detection on NQ-Open

0.753AUCsSHADE

Evaluation Results

Method	Links
SHADE2026.04		0.753	0.678
Hb Hybrid2026.04		0.752	0.676
NumSets2026.04		0.747	0.685
Hb Plugin2026.04		0.743	0.693
NumSets2026.04		0.728	0.679
Hb Hybrid2026.04		0.726	0.679
SHADE2026.04		0.725	0.681
\|S\| Hybrid2026.04		0.723	0.687
\|S\| Hybrid2026.04		0.722	0.686
KLE2026.04		0.715	0.672
Hb Plugin2026.04		0.714	0.684
SHADE2026.04		0.678	0.755
Hb Hybrid2026.04		0.677	0.753
KLE2026.04		0.665	0.687
NumSets2026.04		0.659	0.742
\|S\| Hybrid2026.04		0.655	0.692
Hb Plugin2026.04		0.651	0.735
DSE2026.04		0.628	0.657
DSE2026.04		0.615	0.651
DSE2026.04		0.604	0.659
KLE2026.04		0.588	0.658