List your company

Multi-task Language Understanding on MMLU-Pro (Wall-clock time)

0.84Wall-clock Time (s)DSC
0.76161.29081.822.3492Nov 4, 2025

Evaluation results

MethodLinks
DSCBudget (B)=16, LLM batch size=12025.110.84
CGES-LNS (arith)Budget (B)=16, LLM batch size=12025.111.27
CGES-DeepConf (B10)Budget (B)=16, LLM batch size=1, DeepConf setting=B102025.111.27
CGES-DeepConf (tail)Budget (B)=16, LLM batch size=1, DeepConf setting=tail2025.111.27
ASC-βBudget (B)=16, LLM batch size=12025.111.31
ESCBudget (B)=16, LLM batch size=1, window size w=42025.111.38
CGES-LNS (geom)Budget (B)=16, LLM batch size=12025.111.38
ASC-DirBudget (B)=16, LLM batch size=12025.111.39
CGES-MARSBudget (B)=16, LLM batch size=12025.111.67
ESCBudget (B)=16, LLM batch size=1, window size w=82025.111.97
SCBudget (B)=16, LLM batch size=12025.112.45
CGES-RMBudget (B)=16, LLM batch size=12025.112.8