List your company

Mathematical Reasoning on AMC12 (Pass@1, Pass@3)

66.27Pass@1P-ALIGN
31.58640.590549.59558.5995Jan 15, 2026

Evaluation results

MethodLinks
P-ALIGNBackbone=Qwen3-8B2026.0166.2777.11
UPFT (2025)Backbone=Qwen3-8B2026.0161.7475.38
Zero-ShotBackbone=Qwen3-8B2026.0159.8471.08
SFT (Long-CoT)Backbone=Qwen3-8B2026.0158.8167.04
P-ALIGNBackbone=Qwen2.5-7B-Instruct2026.0149.463.86
SFT (Label)Backbone=Qwen3-8B2026.0148.5159.04
SFT (Long-CoT)Backbone=Qwen2.5-7B-Instruct2026.0148.1956.63
UPFT (2025)Backbone=Qwen2.5-7B-Instruct2026.0147.5859.04
Zero-ShotBackbone=Qwen2.5-7B-Instruct2026.0140.7649.7
SFT (Label)Backbone=Qwen2.5-7B-Instruct2026.0132.9234.62