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

Multi-task Model Merging Evaluation on Multi-task Category-level Evaluation Suite Qwen-3-4B

1.18Average Score

METIS

0.978241.030621.0831.13538Jun 15, 2026
Updated 1mo ago

Evaluation Results

MethodLinks
2026.06
1.181.3180.921.0621.655
2026.06
1.1541.1360.8861.0871.638
2026.06
1.1281.1820.831.0521.603
2026.06
1.1081.1360.8861.0291.534
2026.06
1.08710.8181.0441.569
2026.06
1.0840.6820.9771.0691.638
2026.06
1.06710.751.041.534
2026.06
1.0480.9090.751.0361.517
2026.06
0.9860.8180.6140.9941.5