List your company

Mathematical Reasoning on AIME 2025 (Acc %, Tokens)

86.67AccuracyRAG
3.4725.0746.6768.27Sep 26, 2025

Evaluation results

MethodLinks
RAGBackbone=Qwen3-14B2025.0986.679,910.86
CoT-SCBackbone=Qwen3-4B2025.098042,002.8
CoT-SCBackbone=Qwen3-8B2025.098043,571.8
RoTBackbone=Qwen3-14B2025.098010,187
CoTBackbone=Qwen3-14B2025.0973.339,451.27
RoT + TIBackbone=Qwen3-4B2025.0966.679,421
RAGBackbone=Qwen3-8B2025.0966.6711,777.46
CoT-SCBackbone=Qwen3-14B2025.0966.6739,272.87
RoT + TIBackbone=Qwen3-14B2025.0966.679,767.33
RAGBackbone=Qwen3-4B2025.096011,670.26
RoTBackbone=Qwen3-4B2025.096011,634.6
CoTBackbone=Qwen3-8B2025.096012,770.6
RoTBackbone=Qwen3-8B2025.096012,831.67
CoTBackbone=Qwen3-4B2025.0953.3311,488.33
RoT + TIBackbone=Qwen3-8B2025.0953.3311,279.07
BoTBackbone=Qwen3-8B2025.0933.3359,970.86
CoTBackbone=Qwen3-1.7B2025.0926.6713,111.87
RAGBackbone=Qwen3-1.7B2025.0926.6713,280.8
BoTBackbone=Qwen3-4B2025.0926.6750,194.53
CoT-SCBackbone=Qwen3-1.7B2025.092047,804.8
RoTBackbone=Qwen3-1.7B2025.092012,791.73
RoT + TIBackbone=Qwen3-1.7B2025.09208,568.47
RAGBackbone=Qwen3-0.6B2025.0913.3310,523.86
CoT-SCBackbone=Qwen3-0.6B2025.0913.3344,520.4
BoTBackbone=Qwen3-1.7B2025.0913.3333,657.47
CoTBackbone=Qwen3-0.6B2025.096.6710,004.93
BoTBackbone=Qwen3-0.6B2025.096.6738,792.17
RoTBackbone=Qwen3-0.6B2025.096.6711,994
RoT + TIBackbone=Qwen3-0.6B2025.096.675,277.87