List your company

Mathematical Reasoning on GSM8K (Acc, Cost)

93.21AccuracyCoT
-2.345222.462447.2772.0776Mar 14, 2026

Evaluation results

MethodLinks
CoTModel=Gemma32026.0393.21850
CoTModel=Qwen3-8B2026.0389.09799.7
CoTModel=Llama3.12026.0387.52817
ToTModel=Qwen3-8B2026.033.693,175
ToTModel=Llama3.12026.033.363,033
DSTModel=Qwen3-8B2026.033.35192.3
ToTModel=Gemma32026.032.943,650
DSTModel=Llama3.12026.032.65241
DSTModel=Gemma32026.032.63400
DPTSModel=Llama3.12026.031.722,506
DPTSModel=Qwen3-8B2026.031.532,670
DPTSModel=Gemma32026.031.333,050