List your company

Math on GSM8K (Workspace Metrics)

100True Workspace RatePromptCOS
9.5233.0156.579.99Sep 3, 2025

Evaluation results

MethodLinks
PromptCOSModel=TinyLlama-chat2025.09100210.72
PromptCOSModel=Deepseek-d-qwen2025.09100170.64
PromptCOSModel=Gemma2-it2025.0910070.89
PCGModel=Gemma2-it2025.099850.91
PRModel=Gemma2-it2025.0983790.03
PCGModel=TinyLlama-chat2025.0982120.65
PRModel=TinyLlama-chat2025.0976610.03
PRModel=Deepseek-d-qwen2025.097472-0.02
PC*Model=Gemma2-it2025.095670.45
PC*Model=TinyLlama-chat2025.095420.49
PC*Model=Deepseek-d-qwen2025.094510.42
PCGModel=Deepseek-d-qwen2025.091300.11