Confidence Estimation on MMLU (test) benchmark leaderboard · SOTA2 Research

Confidence Estimation on MMLU (test)

0.07ECE (Smoothed)NWKR

Evaluation Results

Method	Links
NWKR2026.06		0.07	0.9	0.73	0.74	0.79
ACUTE2026.06		0.07	0.91	0.73	0.74	0.79
ACUTE2026.06		0.07	0.91	0.76	0.78	0.82
ACUTE2026.06		0.07	0.91	0.77	0.8	0.83
ACUTE2026.06		0.08	0.91	0.76	0.78	0.82
ACUTE2026.06		0.08	0.91	0.76	0.77	0.81
ACUTE2026.06		0.09	0.9	0.75	0.78	0.81
HRE2026.06		0.11	0.87	0.72	0.71	0.77
Raw Conf2026.06		0.17	0.9	0.71	0.54	0.72