List your company

Multi-domain Evaluation on MMLU, GSM8K, HEval, BBH, MedQA

68Average ScoreCODA
53.9657.60561.2564.895Jul 3, 2026

Evaluation results

MethodLinks
CODABackbone=Qwen-2.5-7B, Data Budget=50%2026.0768
GradNormBackbone=Qwen-2.5-7B, Data Budget=50%2026.0766.8
GREATSBackbone=Qwen-2.5-7B, Data Budget=50%2026.0766.7
PCGradBackbone=Qwen-2.5-7B, Data Budget=50%2026.0766.6
DoReMiBackbone=Qwen-2.5-7B, Data Budget=50%2026.0766.4
Full dataBackbone=Qwen-2.5-7B, Data Budget=100%2026.0766.2
TemperatureBackbone=Qwen-2.5-7B, Data Budget=50%2026.0766
ProportionalBackbone=Qwen-2.5-7B, Data Budget=50%2026.0765.7
Uniform mixingBackbone=Qwen-2.5-7B, Data Budget=50%2026.0765.5
CODABackbone=LLaMA-3.1-8B, Data Budget=50%2026.0757.2
GradNormBackbone=LLaMA-3.1-8B, Data Budget=50%2026.0756
GREATSBackbone=LLaMA-3.1-8B, Data Budget=50%2026.0755.8
PCGradBackbone=LLaMA-3.1-8B, Data Budget=50%2026.0755.7
DoReMiBackbone=LLaMA-3.1-8B, Data Budget=50%2026.0755.5
Full dataBackbone=LLaMA-3.1-8B, Data Budget=100%2026.0755.4
TemperatureBackbone=LLaMA-3.1-8B, Data Budget=50%2026.0755.1
ProportionalBackbone=LLaMA-3.1-8B, Data Budget=50%2026.0754.7
Uniform mixingBackbone=LLaMA-3.1-8B, Data Budget=50%2026.0754.5