Share your thoughts, 1 month free Claude Pro on usSee more
WorkDL logo mark

Multi-task learning evaluation on MM-MergeBench 20 tasks 1.0 (test)

91.37Average Absolute Accuracy

Individual

54.689264.212173.73583.2579Mar 6, 2026
Updated 4mo ago

Evaluation Results

MethodLinks
2026.03
91.37-
2026.03
80.5888.18
2026.03
78.0585.48
2026.03
77.0784.29
2026.03
69.976.71
2026.03
56.1-