List your company

Comprehensive Reasoning on Mathematical and General Reasoning Suite Combined

46.7Overall AccuracyWIST
11.8620.90529.9538.995Mar 22, 2026

Evaluation results

MethodLinks
WISTBackbone=Qwen3-8B-Base2026.0346.7
SPICEBackbone=Qwen3-8B-Base2026.0346
R-ZeroBackbone=Qwen3-8B-Base2026.0345.5
WISTBackbone=Qwen3-4B-Base2026.0343.1
Base ModelBackbone=Qwen3-8B-Base2026.0342.1
SPICEBackbone=Qwen3-4B-Base2026.0341.8
R-ZeroBackbone=Qwen3-4B-Base2026.0340.6
Base ModelBackbone=Qwen3-4B-Base2026.0333.3
WISTBackbone=OctoThinker-8B-Hybrid-Base2026.0332.6
SPICEBackbone=OctoThinker-8B-Hybrid-Base2026.0331.4
R-ZeroBackbone=OctoThinker-8B-Hybrid-Base2026.0330.6
SPICEBackbone=OctoThinker-3B-Hybrid-Base2026.0325.2
WISTBackbone=OctoThinker-3B-Hybrid-Base2026.0324.5
Base ModelBackbone=OctoThinker-8B-Hybrid-Base2026.0322.9
R-ZeroBackbone=OctoThinker-3B-Hybrid-Base2026.0322.5
Base ModelBackbone=OctoThinker-3B-Hybrid-Base2026.0313.2