List your company

Mathematical Problem Solving on MATH 500 (Accuracy, Mean, Drop)

95.66AccuracyBF16
-3.826422.001847.8373.6582May 18, 2026

Evaluation results

MethodLinks
BF16Model=GLM-4.7 FP8, 358B, BPE=16.002026.0595.6677.89—
TurboQuant (no MP)Model=GLM-4.7 FP8, 358B, BPE=3.252026.0595.3978.150.26
Saw-INT4Model=GLM-4.7 FP8, 358B, BPE=4.252026.0595.3277.950.06
OSCARModel=GLM-4.7 FP8, 358B, BPE=2.282026.0594.6678.160.27
BF16Model=Qwen3-4B Thinking-2507, BPE=16.002026.0593.5575.64—
BF16Model=Qwen3-32B, BPE=16.002026.0593.5574.19—
Saw-INT4Model=Qwen3-32B, BPE=4.252026.0593.3174.430.24
Saw-INT4Model=Qwen3-4B Thinking-2507, BPE=4.252026.0593.1973.11-2.53
OSCARModel=Qwen3-4B Thinking-2507, BPE=2.282026.0592.7571.864-3.78
OSCARModel=Qwen3-32B, BPE=2.282026.0592.7574.17-0.02
Saw-INT4Model=Qwen3-8B, BPE=4.252026.0592.6369.97-0.87
BF16Model=Qwen3-8B, BPE=16.002026.0592.5970.84—
OSCARModel=Qwen3-8B, BPE=2.282026.0592.2269.416-1.42
TurboQuant (no MP)Model=Qwen3-32B, BPE=3.252026.0590.671.99-2.2
QuaRot-INT2Model=GLM-4.7 FP8, 358B, BPE=2.252026.0590.475.14-2.75
TurboQuant (no MP)Model=Qwen3-8B, BPE=3.252026.058756.88-13.96
Naive INT2Model=GLM-4.7 FP8, 358B, BPE=2.252026.0586.660.49-17.4
TurboQuant (no MP)Model=Qwen3-4B Thinking-2507, BPE=3.252026.0568.231.74-43.9
QuaRot-INT2Model=Qwen3-8B, BPE=2.252026.0523.1310.14-60.7
QuaRot-INT2Model=Qwen3-32B, BPE=2.252026.0517.67.9-66.29
QuaRot-INT2Model=Qwen3-4B Thinking-2507, BPE=2.252026.055.671.4-74.24
Naive INT2Model=Qwen3-4B Thinking-2507, BPE=2.252026.0500-75.64
Naive INT2Model=Qwen3-8B, BPE=2.252026.0500-70.84
Naive INT2Model=Qwen3-32B, BPE=2.252026.0500-74.19