List your company
ResearchFormerly WizWand

Mathematical Reasoning on AIME 2025 (Latency)

17.07Latency (s)RoT + TI
-32.4724301.9388636.35970.7612Sep 26, 2025
Updated 7mo ago

Evaluation results

MethodLinks
RoT + TIModel=Qwen3-0.6B2025.0917.07
RoT + TIModel=Qwen3-1.7B2025.0998.87
CoTModel=Qwen3-0.6B2025.09116.85
RAGModel=Qwen3-0.6B2025.09127.03
RoTModel=Qwen3-0.6B2025.09129.02
RAGModel=Qwen3-1.7B2025.09156.58
CoTModel=Qwen3-1.7B2025.09163
RoT + TIModel=Qwen3-4B2025.09163.43
RoTModel=Qwen3-1.7B2025.09176.77
CoTModel=Qwen3-4B2025.09195.11
RAGModel=Qwen3-4B2025.09206.86
RoT + TIModel=Qwen3-8B2025.09228.72
RoTModel=Qwen3-4B2025.09245.36
BoTModel=Qwen3-0.6B2025.09270
RoT + TIModel=Qwen3-14B2025.09278.88
RAGModel=Qwen3-8B2025.09285.09
CoTModel=Qwen3-8B2025.09285.25
RoTModel=Qwen3-8B2025.09303.62
RAGModel=Qwen3-14B2025.09311.75
CoTModel=Qwen3-14B2025.09322.19
RoTModel=Qwen3-14B2025.09331.5
BoTModel=Qwen3-1.7B2025.09369.73
CoT-SCModel=Qwen3-0.6B2025.09542.92
BoTModel=Qwen3-4B2025.09560.21
CoT-SCModel=Qwen3-1.7B2025.09672.93
BoTModel=Qwen3-8B2025.09694.16
CoT-SCModel=Qwen3-4B2025.09870.46
CoT-SCModel=Qwen3-8B2025.091,109.18
CoT-SCModel=Qwen3-14B2025.091,255.63