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

Mathematical Reasoning on Math Category Average (test)

51.5Accuracy

RLVR

Updated 1mo ago

Evaluation Results

Method	Links
RLVR 2026.03		51.5	0.63	0.46	0.45
RLVR+QuestionProb 2026.03		50.23	0.32	0.47	0.47
RLVR+Assessor Model 2026.03		50.23	0.83	0.24	0.19
RLVR+Probe 2026.03		50.23	0.84	0.12	0.1
CoCA 2026.03		47.92	0.71	0.1	0.09
RLVR+QuestionProb 2026.03		46.53	0.42	0.44	0.45
RLVR+Assessor Model 2026.03		46.53	0.83	0.16	0.14
RLVR+Probe 2026.03		46.53	0.84	0.11	0.11
CoCA 2026.03		44.85	0.88	0.09	0.09
Instruct Model 2026.03		43.73	0.61	0.52	0.5
RLVR 2026.03		42.86	0.34	0.48	0.47
RLVR+QuestionProb 2026.03		39.4	0.44	0.36	0.37
RLVR+Assessor Model 2026.03		39.4	0.72	0.16	0.14
RLVR+Probe 2026.03		39.4	0.62	0.1	0.12
Instruct Model 2026.03		37.54	0.54	0.54	0.52
CoCA 2026.03		36.85	0.82	0.09	0.1
RLVR 2026.03		30.89	0.52	0.62	0.61
Instruct Model 2026.03		13.79	0.52	0.77	0.73