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

Answer Verification on MATH

0.9128AUROC

MSV 64

Updated 1mo ago

Evaluation Results

Method	Links
MSV 64 2026.03		0.9128	0.1147
MSV 16 2026.03		0.9062	0.1255
MSV 64 2026.03		0.9049	0.1261
MSV 16 2026.03		0.9037	0.1104
MSV 4 2026.03		0.8945	0.1348
MSV 4 2026.03		0.8917	0.1324
Probe 2026.03		0.879	0.1346
Probe + WV 4 2026.03		0.8769	0.1563
Probe + WV 16 2026.03		0.8769	0.1563
Probe + WV 64 2026.03		0.8769	0.1563
Probe 2026.03		0.869	0.1592
MSV 1 + WV 4 2026.03		0.8567	0.1629
MSV 1 + WV 16 2026.03		0.8567	0.1629
MSV 1 + WV 64 2026.03		0.8567	0.1629
MSV 1 2026.03		0.8566	0.146
MSV 1 2026.03		0.8443	0.1601
MSV 1 + WV 64 2026.03		0.8364	0.138
Probe + WV 64 2026.03		0.8346	0.137
Probe(EOS) 2025.08		0.834	-
Probe(AVG) 2025.08		0.833	-
Prob(Exact) 2025.08		0.827	-
Probe + WV 16 2026.03		0.8228	0.1416
MSV 1 + WV 16 2026.03		0.8228	0.1425
Probe(EU) 2025.08		0.794	-
Probe + WV 4 2026.03		0.793	0.1649
MSV 1 + WV 4 2026.03		0.792	0.1656
LogTokU 2025.08		0.791	-
Probe(EOS) 2025.08		0.79	-
Probe(AVG) 2025.08		0.786	-
P(true) 2025.08		0.779	-
Prob(Exact) 2025.08		0.773	-
LogProb 2025.08		0.757	-
Probe(EOS) 2025.08		0.703	-
Probe(AVG) 2025.08		0.699	-
LogProb 2025.08		0.683	-
Probe(EU) 2025.08		0.669	-
LogTokU 2025.08		0.666	-
P(true) 2025.08		0.664	-
Probe(EU) 2025.08		0.646	-
P(true) 2025.08		0.635	-
LogProb 2025.08		0.635	-
Prob(Exact) 2025.08		0.627	-
LogTokU 2025.08		0.565	-