Share your thoughts, 1 month free Claude Pro on usSee more

Process-outcome alignment on PRIME Math

92.44Outcome Score

Claude-Sonnet-4.5-thinking

Updated 1mo ago

Evaluation Results

Method	Links
Claude-Sonnet-4.5-thinking 2026.02		92.44	81
Gemini-2.5-Pro-thinking 2026.02		92.11	86.11
GPT-5.2-thinking 2026.02		92.11	86
Kimi-K2-thinking 2026.02		91.78	82.56
Claude-Opus-4-thinking 2026.02		91.59	80.89
Grok-4-thinking 2026.02		91.33	85.78
GPT-OSS-120B 2026.02		91.22	78.33
GPT-5-thinking 2026.02		91.11	86.67
Claude-Opus-4.5 2026.02		90.89	75.67
GPT-5.2-chat 2026.02		90.78	81.89
Claude-Sonnet-4-thinking 2026.02		90.78	81
DeepSeek-V3.2-thinking 2026.02		90.67	83.89
Gemini-3.0-pro-thinking 2026.02		90.56	81.67
Qwen3-30B-A3B 2026.02		90.44	79.67
GLM-4.6 2026.02		90.22	81
Claude-Sonnet-4.5 2026.02		90.22	74
Seed-1.6-thinking 2026.02		90	74.78
GPT-OSS-20B 2026.02		89.67	75.56
Claude-Sonnet-4 2026.02		89.44	74.78
Claude-Opus-4 2026.02		89.33	75.33
Qwen3-4B-thinking-2507 2026.02		89.22	76.22
Claude-Opus-4.5-thinking 2026.02		88.89	79.33
CompassVerifier-32B 2026.02		88.33	67.33
GPT-5-chat 2026.02		88.11	71.11
Qwen3-8B 2026.02		87.78	72.89
DeepSeek-V3.2 2026.02		87.56	72.44
SCI-Verifier-4B 2026.02		87.56	68.78
xVerify-9B-C 2026.02		87.11	67.44
Qwen3-4B 2026.02		86.33	69.56
Tencent-Qwen2.5-7B-RLVR 2026.02		86.22	66.33
Phi-4 2026.02		81.44	60
Qwen3-4B-Instruct-2507 2026.02		81.22	64.78
General-Verifier 2026.02		76.78	57.78
Math-Verify 2026.02		65.56	63.67