List your company

Mathematical Reasoning on Mathematical Reasoning Suite Aggregate

66.71Average ScoreMIPU
50.22654.505558.78563.0645Jun 28, 2026

Evaluation results

MethodLinks
MIPUModel=Qwen3-4B, Rollout Quantization=FP82026.0666.71—
LR-decayModel=Qwen3-4B, Rollout Quantization=FP82026.0665.66—
GRPO BaselineModel=Qwen3-4B, Rollout Quantization=FP82026.0664.42—
MISModel=Qwen3-4B, Rollout Quantization=FP82026.0663.42—
MIPUModel=Qwen3-1.7B, Rollout Quantization=FP82026.0653.97—
LR-decayModel=Qwen3-1.7B, Rollout Quantization=FP82026.0652.23—
MISModel=Qwen3-1.7B, Rollout Quantization=FP82026.0651.73—
GRPO BaselineModel=Qwen3-1.7B, Rollout Quantization=FP82026.0650.86—