List your company

Mathematical Reasoning on AMC 2023 (Accuracy and Gen. Cost)

93.5AccuracyZIP-RC sampling
57.166.557685.45Dec 1, 2025

Evaluation results

MethodLinks
ZIP-RC samplingModel=LFM2-1.2B2025.1293.51.35
Weighted BoN Self-evalModel=LFM2-1.2B2025.1291.81.6
ZIP-RC samplingModel=Qwen3-1.7B2025.1290.91.43
Majority VotingModel=LFM2-1.2B2025.1290.61.6
ZIP-RC reward pruneModel=LFM2-1.2B2025.1290.21.49
MV length-pruneModel=LFM2-1.2B2025.1289.81.7
Weighted BoN Self-evalModel=Qwen3-1.7B2025.1289.11.4
Weighted BoN ext. RMModel=LFM2-1.2B2025.12891.53
Majority VotingModel=Qwen3-1.7B2025.1287.91.4
Weighted BoN ext. RMModel=Qwen3-1.7B2025.1286.51.43
ZIP-RC reward pruneModel=Qwen3-1.7B2025.12861.33
ZIP-RC samplingModel=LFM2-350M2025.1283.91.49
Weighted BoN Self-evalModel=LFM2-350M2025.1277.61.7
MV length-pruneModel=LFM2-350M2025.1274.81.66
Majority VotingModel=LFM2-350M2025.1274.51.7
Weighted BoN ext. RMModel=LFM2-350M2025.1273.41.59
ZIP-RC reward pruneModel=LFM2-350M2025.1269.71.27
MV length-pruneModel=Qwen3-1.7B2025.1258.51.46