List your company

Mathematical Reasoning on (MATH-500, Olympiad, Minerva, GSM8K, AMC, AIME24) (test)

58.12Overall ScorePretrainZero
13.847225.341136.83548.3289Dec 3, 2025

Evaluation results

MethodLinks
PretrainZeroModel Series=Qwen3-30B-A3B-MoE-Base2025.1258.1281.743.441.294.470.6217.4
PretrainZeroModel Series=Qwen3-8B-Base2025.1257.7281.942.543.493.56520
Random RLPTModel Series=Qwen3-8B-Base2025.1255.0879.242.73993.862.513.3
PretrainZeroRLPT steps=2000, RLVR steps=4002025.1253.7778.84339.79354.8413.3
PretrainZeroModel Series=Qwen3-4B-Base2025.1253.1379.142.733.892.956.9513.3
Random RLPTModel Series=Qwen3-30B-A3B-MoE-Base2025.1252.6279.241.238.682.459.7714.58
Base ModelModel Series=Qwen3-30B-A3B-MoE-Base2025.1252.4974.74322.891.166.9516.36
PretrainZeroRLPT steps=1000, RLVR steps=4002025.1251.84774332.492.55511.13
Base ModelRLPT steps=0, RLVR steps=4002025.1250.9675.741.831.691.352.0313.3
Random RLPTRLPT steps=2000, RLVR steps=4002025.1247.8774.838.522.187.554.310
Random RLPTModel Series=Qwen3-4B-Base2025.1247.8774.838.522.187.554.310
Base ModelModel Series=Qwen3-8B-Base2025.1247.4870.135.325.491.552.5810
Base ModelModel Series=Qwen3-4B-Base2025.1242.5373.337.322.186.336.170
PretrainZeroModel Series=SmolLM3-3B-Base2025.1238.0362.625.622.183.727.56.7
Random RLPTModel Series=SmolLM3-3B-Base2025.1235.955921.520.282.532.50
Base ModelModel Series=SmolLM3-3B-Base2025.1232.3153.820.41481.222.811.65
Continue PTModel Series=Qwen3-8B-Base2025.1227.7842.716.911.855.333.286.7
Continue PTModel Series=Qwen3-4B-Base2025.1224.653813.61167153.3
Supervised FTModel Series=Qwen3-8B-Base2025.1219.2330.511.715.432.8250
Supervised FTModel Series=Qwen3-4B-Base2025.1215.5528.58.114.327.4150