List your company

Mathematical Reasoning on GSM8K (P@1, P@4)

86.68P@1ICT
10.96830.62450.2869.936Jun 18, 2026

Evaluation results

MethodLinks
ICTModel Size=Qwen2.5-7B2026.0686.6894.47
STAPOModel Size=Qwen2.5-7B2026.0685.2493.93
GRPOModel Size=Qwen2.5-7B2026.0684.794.39
20-EntropyModel Size=Qwen2.5-7B2026.0670.3291.81
BaseModel Size=Qwen2.5-7B2026.0667.6690.98
ICTModel Size=Qwen2.5-1.5B2026.0666.2385.31
STAPOModel Size=Qwen2.5-1.5B2026.0662.3180.15
GRPOModel Size=Qwen2.5-1.5B2026.0661.5181.71
20-EntropyModel Size=Qwen2.5-1.5B2026.0659.582.12
BaseModel Size=Qwen2.5-1.5B2026.0643.177.29
GRPOModel Size=Qwen2.5-0.5B2026.0642.8763.15
STAPOModel Size=Qwen2.5-0.5B2026.0640.8864.5
ICTModel Size=Qwen2.5-0.5B2026.0640.2765.4
20-EntropyModel Size=Qwen2.5-0.5B2026.0630.3850.5
BaseModel Size=Qwen2.5-0.5B2026.0613.8838.62