List your company

Tool Use on T-Eval

80.7Overall ScorePROVE
66.34870.07473.877.526Jun 2, 2026

Evaluation results

MethodLinks
PROVEModel=Qwen3-8B2026.0680.794.865.26392.781.886.5—
Qwen3-8B + PROVETraining Steps=350 GRPO steps2026.0680.7——————1.9
BaseModel=Qwen3-8B2026.0678.898.864.764.28877.879.2—
PROVEModel=Qwen3-4B2026.0678.298.568.763.790.678.768.8—
Qwen3-4B + PROVETraining Steps=350 GRPO steps2026.0678.2——————1.5
PROVEModel=Granite-4.1-8B2026.0677.994.769.763.791.776.970.8—
Granite-4.1-8B + PROVETraining Steps=350 GRPO steps2026.0677.9——————4.4
BaseModel=Qwen3-4B2026.0676.795.566.462.887.577.370.8—
BaseModel=Granite-4.1-8B2026.0673.593.568.263.4877455.2—
PROVEModel=Qwen2.5-7B2026.0673.481.860.463.18869.777.1—
Qwen2.5-7B + PROVETraining Steps=350 GRPO steps2026.0673.4——————6.5
BaseModel=Qwen2.5-7B2026.0666.973.454.157.779.262.275—