List your company

Mathematical Reasoning on GSM8K (Accuracy Deg %, BERTScore)

2.15Accuracy Deg %PCG
-1.0012-0.18310.6351.4531Sep 3, 2025

Evaluation results

MethodLinks
PCGModel=Deepseek-d-qwen2025.092.150.43
PCGModel=TinyLlama-chat2025.092.080.51
PCGModel=Gemma2-it2025.091.730.5
PC*Model=Gemma2-it2025.090.560.61
PC*Model=Deepseek-d-qwen2025.090.450.77
PC*Model=TinyLlama-chat2025.090.220.64
PromptCOSModel=Gemma2-it2025.090.010.68
PromptCOSModel=TinyLlama-chat2025.09-0.080.63
PromptCOSModel=Deepseek-d-qwen2025.09-0.880.72