List your company

General LLM Capability Evaluation on MMLU, AlpacaEval, GSM8K, MATH, HumanEval

71.86MMLUBFPO
50.082455.736261.3967.0438Jun 1, 2026

Evaluation results

MethodLinks
BFPOBackbone=Qwen2.5-7B, Temperature=1.02026.0671.8694.9170.6625.682.9369.19
BFPOBase Model=Qwen2.5-7B-Instruct, Temperature=02026.0671.8694.9171.4925.683.5469.48
NSPOBackbone=Qwen2.5-7B, Temperature=1.02026.0671.7695.2869.3726.679.8868.58
NSPOBase Model=Qwen2.5-7B-Instruct, Temperature=02026.0671.7695.2870.5123.882.9368.86
Qwen2.5-7B-InstructBackbone=Qwen2.5-7B, Temperature=1.02026.0671.795.9670.1225.679.2668.53
Qwen2.5-7B-InstructTemperature=02026.0671.795.9670.3524.881.768.9
SafeSteerBackbone=Qwen2.5-7B, Temperature=1.02026.0671.6393.9870.5825.476.8367.68
SafeSteerBase Model=Qwen2.5-7B-Instruct, Temperature=02026.0671.6393.9872.7125.282.3269.17
DPO-MixBackbone=Qwen2.5-7B, Temperature=1.02026.0671.6154.0475.5126.475.6160.63
DPO-MixBase Model=Qwen2.5-7B-Instruct, Temperature=02026.0671.6159.1377.125.673.7861.44
SafeSteerBackbone=Qwen2-1.5B, Temperature=1.02026.0670.6794.7872.7827.889.6371.13
SafeSteerBase Model=Qwen3-4B-Instruct, Temperature=02026.0670.6794.7873.4727.290.8571.39
NSPOBackbone=Qwen2-1.5B, Temperature=1.02026.0670.6595.7873.6926.487.870.86
NSPOBase Model=Qwen3-4B-Instruct, Temperature=02026.0670.6595.7873.2426.690.8571.42
Qwen2-1.5B-InstructBackbone=Qwen2-1.5B, Temperature=1.02026.0670.6196.0274.0726.289.6371.31
DPO-MixBackbone=Qwen2-1.5B, Temperature=1.02026.0670.6158.4974.2227.289.6364.03
Qwen3-4B-InstructTemperature=02026.0670.6196.0273.0826.692.0771.68
DPO-MixBase Model=Qwen3-4B-Instruct, Temperature=02026.0670.6168.7675.8929.490.8567.1
BFPOBackbone=Qwen2-1.5B, Temperature=1.02026.0670.695.2873.012690.8571.15
BFPOBase Model=Qwen3-4B-Instruct, Temperature=02026.0670.695.2873.312890.2471.49
MoCANBackbone=Qwen2.5-7B, Temperature=1.02026.0670.1894.1668.6825.482.9268.27
MoCANBase Model=Qwen2.5-7B-Instruct, Temperature=02026.0670.1894.1668.9124.284.7668.44
W-DOORBackbone=Qwen2.5-7B, Temperature=1.02026.066884.7661.7126.230.4954.23
W-DOORBase Model=Qwen2.5-7B-Instruct, Temperature=02026.066884.7663.8427.626.8354.21
W-DOORBackbone=Qwen2-1.5B, Temperature=1.02026.0665.2690.9940.1814.224.3947
W-DOORBase Model=Qwen3-4B-Instruct, Temperature=02026.0665.2690.9942.291427.4448
Llama-3-8B-InstructBase Model=Llama-3-8B-Instruct, Temperature=02026.0664.7296.1566.343060.9863.64
Llama-3-8B-InstructBase Model=Llama-3-8B, Temperature=1.02026.0664.7296.1561.5224.454.2760.21
BFPOBase Model=Llama-3-8B-Instruct, Temperature=02026.0664.6695.963.7629.860.9863.02
BFPOBase Model=Llama-3-8B, Temperature=1.02026.0664.6695.960.7327.456.7161.08
SafeSteerBase Model=Llama-3-8B-Instruct, Temperature=02026.0664.6395.2765.043060.9863.18
SafeSteerBase Model=Llama-3-8B, Temperature=1.02026.0664.6395.2763.682353.6660.05
NSPOBase Model=Llama-3-8B-Instruct, Temperature=02026.0664.5895.8464.8230.459.7663.08
NSPOBase Model=Llama-3-8B, Temperature=1.02026.0664.5895.8461.9412.857.9358.62
DPO-MixBase Model=Llama-3-8B-Instruct, Temperature=0, Mixing Ratio=50/502026.0664.479.2573.1628.860.3761.2
DPO-MixBase Model=Llama-3-8B, Temperature=1.02026.0664.476.5271.4924.659.7659.35
MoCANBase Model=Llama-3-8B-Instruct, Temperature=02026.0663.6694.3571.1126.257.9362.65
MoCANBackbone=Qwen2-1.5B, Temperature=1.02026.0662.9895.1472.6327.289.0269.39
MoCANBase Model=Qwen3-4B-Instruct, Temperature=02026.0662.9895.1472.1826.290.2469.35
Llama-3.2-3B-InstructBase Model=Llama-3.2-3B-Instruct, Temperature=02026.0662.1992.872.7837.653.6763.61
Llama-3.2-3B-InstructBase Model=Llama-3.2-3B, Temperature=1.02026.0662.1992.863.0835.446.9560.08
BFPOBase Model=Llama-3.2-3B-Instruct, Temperature=02026.0662.1890.9972.3337.452.4463.07
NSPOBase Model=Llama-3.2-3B-Instruct, Temperature=02026.0662.1891.4971.953852.4463.21
BFPOBase Model=Llama-3.2-3B, Temperature=1.02026.0662.1890.9960.7334.853.0560.35
NSPOBase Model=Llama-3.2-3B, Temperature=1.02026.0662.1891.4963.1532.249.3959.68
DPO-MixBase Model=Llama-3.2-3B-Instruct, Temperature=0, Mixing Ratio=50/502026.0662.0279.7574.6837.454.2761.62
DPO-MixBase Model=Llama-3.2-3B, Temperature=1.02026.0662.0263.6962.5532.648.7853.93
SafeSteerBase Model=Llama-3.2-3B-Instruct, Temperature=02026.0661.6490.872.0236.853.0462.86
SafeSteerBase Model=Llama-3.2-3B, Temperature=1.02026.0661.1690.862.1623.840.2455.63
W-DOORBase Model=Llama-3-8B-Instruct, Temperature=02026.0659.8184.161.4413.230.4937.82
W-DOORBase Model=Llama-3-8B, Temperature=1.02026.0659.8184.161.4413.232.3238.19
MoCANBase Model=Llama-3.2-3B-Instruct, Temperature=02026.0659.7687.4454.0633.454.2757.79
W-DOORBase Model=Llama-3.2-3B-Instruct, Temperature=02026.0655.8379.323.560.431.134.04
W-DOORBase Model=Llama-3.2-3B, Temperature=1.02026.0655.8379.322.431421.3434.58
MoCANBase Model=Llama-3-8B, Temperature=1.02026.0654.5494.3562.722.656.158.06
MoCANBase Model=Llama-3.2-3B, Temperature=1.02026.0650.9287.4441.6922.240.8548.62