List your company

Mathematical Reasoning on MATH-500 (Accuracy and Token Count)

95.52AccuracyCoT
-2.312823.086148.48573.8839Dec 5, 2025Dec 21, 2025Jan 7, 2026Jan 23, 2026Feb 9, 2026Feb 25, 2026Mar 14, 2026

Evaluation results

MethodLinks
CoTModel=Gemma32026.0395.52———3,107
CoTModel=Llama3.12026.0393.55———2,783
CoTModel=Qwen3-8B2026.0392.05———2,680
LYNXBackbone=QwQ-32B, Conf=0.97, Temperature=0.02025.1291.40.443,379.23-14.4—
BaselineBackbone=QwQ-32B, Temperature=0.02025.1291—3,945.49——
LYNXBackbone=QwQ-32B, Conf=0.95, Temperature=0.02025.1290.8-0.223,060.71-22.4—
DEERBackbone=QwQ-32B, Temperature=0.02025.1290.6-0.443,072.83-22.1—
LYNXBackbone=QwQ-32B, Conf=0.90, Temperature=0.02025.1289-2.22,816.04-28.6—
LYNXBackbone=QwQ-32B, Conf=0.80, Temperature=0.02025.1287-4.42,510.93-36.4—
LYNXBackbone=QwQ-32B, Conf=0.70, Temperature=0.02025.1279.8-12.311,968.3-50.1—
LYNXBackbone=DeepSeek-R1-Distill-Qwen-1.5B, Conf=0.95, Temperature=0.02025.1276.212.063,930.17-35.5—
LYNXBackbone=DeepSeek-R1-Distill-Qwen-1.5B, Conf=0.90, Temperature=0.02025.1275.811.473,212.87-47.3—
LYNXBackbone=DeepSeek-R1-Distill-Qwen-1.5B, Conf=0.97, Temperature=0.02025.1274.69.714,511.43-25.9—
ToNBackbone=QwQ-32B, Temperature=0.02025.1273.2-19.562,390.42-39.4—
LYNXBackbone=DeepSeek-R1-Distill-Qwen-1.5B, Conf=0.80, Temperature=0.02025.1270.23.242,427.41-60.1—
BaselineBackbone=DeepSeek-R1-Distill-Qwen-1.5B, Temperature=0.02025.1268—6,090.96——
DEERBackbone=DeepSeek-R1-Distill-Qwen-1.5B, Temperature=0.02025.1266.8-1.762,446.44-59.8—
LYNXBackbone=DeepSeek-R1-Distill-Qwen-1.5B, Conf=0.70, Temperature=0.02025.1260.2-11.471,669.69-72.6—
ToNBackbone=DeepSeek-R1-Distill-Qwen-1.5B, Temperature=0.02025.1254.2-20.293,438.84-43.5—
DPTSModel=Qwen3-8B2026.032.5———2,699
ToTModel=Llama3.12026.032.16———3,869
ToTModel=Qwen3-8B2026.032.15———3,488
DSTModel=Qwen3-8B2026.032.12———762
ToTModel=Gemma32026.032.03———4,358
DPTSModel=Llama3.12026.031.87———2,924
DSTModel=Llama3.12026.031.73———774
DSTModel=Gemma32026.031.58———1,045
DPTSModel=Gemma32026.031.45———3,359