Code Generation on Code Category Average (test) benchmark leaderboard · SOTA2 Research

Code Generation on Code Category Average (test)

77.43AccuracyInstruct Model

Evaluation Results

Method	Links
Instruct Model2026.03		77.43	0.57	0.22	0.23
CoCA2026.03		76.6	0.72	0.16	0.19
RLVR+QuestionProb2026.03		76.58	0.47	0.7	0.68
RLVR+Assessor Model2026.03		76.58	0.55	0.29	0.3
RLVR+Probe2026.03		76.58	0.62	0.71	0.68
RLVR2026.03		76.1	0.55	0.23	0.23
RLVR+QuestionProb2026.03		69.03	0.42	0.65	0.63
RLVR+Assessor Model2026.03		69.03	0.64	0.29	0.33
RLVR+Probe2026.03		69.03	0.59	0.37	0.39
CoCA2026.03		67.38	0.67	0.28	0.3
Instruct Model2026.03		63.78	0.52	0.3	0.33
RLVR+QuestionProb2026.03		57.15	0.46	0.48	0.48
RLVR+Assessor Model2026.03		57.15	0.63	0.25	0.35
RLVR+Probe2026.03		57.15	0.64	0.27	0.34
Instruct Model2026.03		55.1	0.53	0.41	0.42
CoCA2026.03		46.72	0.72	0.09	0.22
RLVR2026.03		39	0.58	0.56	0.55
RLVR2026.03		34.58	0.48	0.56	0.54