List your company

Instruction-Tuning Performance Evaluation on BBH, GPQA, IFEval, Math, MMLU-Pro, and MUSR

52.17BBH ScoreTaskPGM (PMI)
48.966849.798450.6351.4616Jul 16, 2025

Evaluation results

MethodLinks
TaskPGM (PMI)Model=Qwen2-7B, Sample Size=50K2025.0752.1732.839.7321.8342.1747.3
TaskPGM (JSD)Model=Qwen2-7B, Sample Size=50K2025.0752.1633.144719.9641.9146.09
TaskPGM (PMI)Model=Qwen2-7B, Sample Size=25K2025.0752.0233.4139.0920.9642.5147.01
LESSModel=Qwen2-7B, Sample Size=25K2025.0751.7330.232.9720.0242.335.98
LESSModel=Qwen2-7B, Sample Size=50K2025.0751.3629.735.4919.2641.8135.85
UniformModel=Qwen2-7B, Sample Size=50K2025.0751.3332.334.2920.834246.16
RandomModel=Qwen2-7B, Sample Size=50K2025.0751.2832.6337.4119.1841.9846.03
EPMModel=Qwen2-7B, Sample Size=50K2025.0751.1732.7234.1720.0241.9545.37
TaskPGM (JSD)Model=Qwen2-7B, Sample Size=25K2025.0750.9631.841.9721.1541.9845.94
UniformModel=Qwen2-7B, Sample Size=25K2025.0750.1333.1429.2620.8541.6146.83
EPMModel=Qwen2-7B, Sample Size=25K2025.0749.733.1430.9420.6841.4645.37
RandomModel=Qwen2-7B, Sample Size=25K2025.0749.0931.8832.8518.8141.5746.03