Post-deployment performance monitoring on PACS Style Shift benchmark leaderboard · SOTA2 Research

Post-deployment performance monitoring on PACS Style Shift

0.913R2DDBglobal

Evaluation Results

Method	Links
DDBglobal2026.04		0.913	0.921	0.783
CSS2026.04		0.912	0.983	0.944
DDBdeep2026.04		0.891	0.908	0.767
DDBout2026.04		0.862	0.897	0.731
ID Accuracy2026.04		0.765	0.878	0.72
CSS2026.04		0.759	0.883	0.722
AC2026.04		0.646	0.755	0.582
ATC2026.04		0.645	0.617	0.444
ANE2026.04		0.608	0.728	0.554
ATC2026.04		0.555	0.72	0.574
alpha-ReQ2026.04		0.495	0.641	0.474
CSS2026.04		0.383	0.483	0.5
RANKME2026.04		0.379	0.386	0.266
MDE2026.04		0.345	0.65	0.478
CSS2026.04		0.339	0.45	0.278
α-ReQ2026.04		0.325	0.65	0.5
ANE2026.04		0.105	0.2	0.222
RANKME2026.04		0.1	0.6	0.5
CSS2026.04		0.092	0.183	0.167
Sharpness2026.04		0.048	0.075	0.037
AC2026.04		0.031	0.117	0.111
CSS2026.04		0.028	0.5	0.389
MDE2026.04		0.002	0.217	0.222