List your company

Mathematical Reasoning on GSM8K (Acc, Len)

92.95AccuracyLCPO
80.407683.663886.9290.1762Aug 13, 2025

Evaluation results

MethodLinks
LCPOBackbone=DeepSeek-R1-Distill-Qwen-7B2025.0892.95796
L1-MaxBackbone=DeepSeek-R1-Distill-Qwen-7B2025.0892.421,640
L1-ExactBackbone=DeepSeek-R1-Distill-Qwen-7B2025.0892.343,320
OriginalBackbone=DeepSeek-R1-Distill-Qwen-7B2025.0891.811,677
DASTBackbone=DeepSeek-R1-Distill-Qwen-7B2025.0891.211,092
TrEffBackbone=DeepSeek-R1-Distill-Qwen-7B2025.0889.39357
L1-MaxBackbone=DeepSeek-R1-Distill-Qwen-1.5B2025.0888.323,141
CoDBackbone=DeepSeek-R1-Distill-Qwen-7B2025.0888.32416
L1-ExactBackbone=DeepSeek-R1-Distill-Qwen-1.5B2025.0887.952,986
OriginalBackbone=DeepSeek-R1-Distill-Qwen-1.5B2025.0886.282,457
LCPOBackbone=DeepSeek-R1-Distill-Qwen-1.5B2025.0885.82946
TrEffBackbone=DeepSeek-R1-Distill-Qwen-1.5B2025.0883.24612
CoDBackbone=DeepSeek-R1-Distill-Qwen-1.5B2025.0880.891,488