List your company

Multi-task Language Understanding on MMLU Pro (Accuracy and #Calls)

62.1AccuracySC
58.14859.17460.261.226Nov 4, 2025

Evaluation results

MethodLinks
SCBudget (B)=162025.1162.116
ESCBudget (B)=16, Window size (w)=82025.1162.112.87
ESCBudget (B)=16, Window size (w)=42025.11628.99
ASC-βBudget (B)=162025.11628.55
ASC-DirBudget (B)=162025.11629.07
CGES-MARSBudget (B)=162025.11628.8
CGES-LNS (arith)Budget (B)=16, Aggregation function=arith2025.1161.98.24
CGES-LNS (geom)Budget (B)=16, Aggregation function=geom2025.1161.98.95
CGES-DeepConf (B10)Budget (B)=16, DeepConf variant=B102025.1161.98.26
CGES-DeepConf (tail)Budget (B)=16, DeepConf variant=tail2025.1161.98.23
DSCBudget (B)=162025.1160.95.18
CGES-RMBudget (B)=162025.1158.312.31