List your company

Reasoning on AIME24, AIME25, HMMT25 Average

68.6AccuracySC@64
18.78431.71744.6557.583Feb 3, 2026

Evaluation results

MethodLinks
SC@64Base Model=Qwen3-8B, Type=Parallel2026.0268.631.91,100.7
ASCBase Model=Qwen3-8B, Type=Seq.2026.0268.6399.1399.1
ESCBase Model=Qwen3-8B, Type=Hybrid2026.0268.6133.9771.5
Parallel-ProbeBase Model=Qwen3-8B, Type=Parallel2026.0268.521.6824.9
SC@64Base Model=Qwen3-4B, Type=Parallel2026.0266.831.21,047.7
ESCBase Model=Qwen3-4B, Type=Hybrid2026.0266.8136.6770.8
Parallel-ProbeBase Model=Qwen3-4B, Type=Parallel2026.0266.820.9789
ASCBase Model=Qwen3-4B, Type=Seq.2026.0266.7342.2342.2
SC@64 + SACBase Model=Qwen3-4B, Type=Parallel2026.0265.126.6880.2
SC@64 + SACBase Model=Qwen3-8B, Type=Parallel2026.0263.227.4889.5
SC@64Base Model=Qwen3-1.7B, Type=Parallel2026.024731.31,070.9
ASCBase Model=Qwen3-1.7B, Type=Seq.2026.0247556.6556.6
ESCBase Model=Qwen3-1.7B, Type=Hybrid2026.0247168.6945.7
Parallel-ProbeBase Model=Qwen3-1.7B, Type=Parallel2026.0245.121.5794.8
SC@64 + SACBase Model=Qwen3-1.7B, Type=Parallel2026.024226.9888.8
Parallel-ProbeBase Model=Qwen3-0.6B, Type=Parallel2026.0223.320.3735.3
SC@64Base Model=Qwen3-0.6B, Type=Parallel2026.0222.831.6945.7
ASCBase Model=Qwen3-0.6B, Type=Seq.2026.0222.8680680
ESCBase Model=Qwen3-0.6B, Type=Hybrid2026.0222.8181.4926.5
SC@64 + SACBase Model=Qwen3-0.6B, Type=Parallel2026.0220.726.8816.1