Question Answering on MMLU (AUC, F95, AUP) benchmark leaderboard · SOTA2 Research

Question Answering on MMLU (AUC, F95, AUP)

64.07AUCSIVR (Ours)

Evaluation Results

Method	Links
SIVR (Ours)2026.04		64.07	91.3	88.6
SIVR2026.04		64.07	91.3	88.6
SIVR2026.04		55.54	91.46	77.69
SIVR2026.04		55.54	91.46	77.69
TAD2026.04		54.89	94.74	81.29
SAPLMA2026.04		54.66	90.28	80.01
SAPLMA2026.04		54.64	97.06	74.8
Temp-Scaled2026.04		52.86	94.69	76.48
Lookback Lens2026.04		52.66	92.54	77.79
P(True)2026.04		52.63	93.53	74.28
Max Prob2026.04		52.56	—	76.37
Perplexity2026.04		51.9	94.97	76.33
Entropy2026.04		50.96	94.13	76.23
SIVR2026.04		50.89	84.42	74.97
SIVR2026.04		50.89	84.42	74.97
SE2026.04		50.56	96.83	71.36
CoE-C2026.04		50.53	93.45	70.47
SAPLMA2026.04		50.49	95	75.42
CoE-C2026.04		50.42	94.53	73.1
TAD2026.04		50.38	95.89	75.34
TAD2026.04		50.19	91.76	71.39
Temp-Scaled2026.04		50.02	97.02	70.17
CoE-R2026.04		49.87	95.52	72.66
Max Prob2026.04		49.79	97.02	70.02
Perplexity2026.04		49.67	97.02	69.66
SAR2026.04		49.54	92.06	69.51
SATMD + MSP2026.04		49.46	97.22	76.97
SATMD + MSP2026.04		49.46	97.22	76.97
SAR2026.04		49.34	95.83	76.16
Energy2026.04		49.25	93.85	75.35
Entropy2026.04		49.1	96.43	69.16
P(True)2026.04		48.85	95.53	74.96
CoE-R2026.04		48.34	93.45	69.83
SE2026.04		48.03	87.5	74.23
Entropy2026.04		48	96.02	71.72
P(True)2026.04		47.98	96.43	68.41
SATMD + MSP2026.04		47.91	96.43	73.15
SATMD + MSP2026.04		47.91	96.43	73.15
CoE-R2026.04		47.88	96.37	74.33
SE2026.04		47.82	92.31	72.09
Perplexity2026.04		47.76	96.52	71.57
Max Prob2026.04		47.67	96.27	71.42
Temp-Scaled2026.04		47.66	96.27	71.38
Energy2026.04		47.48	92.86	67.25
Energy2026.04		47.33	96.02	72.44
Lookback Lens2026.04		47.14	96.1	69.58
SATMD + MSP2026.04		47.03	94.25	68.08
SATMD + MSP2026.04		47.03	94.25	68.08
CoE-C2026.04		46.2	96.93	73.75
Lookback Lens2026.04		46.13	93.44	72.04
SAR2026.04		45.39	98.08	73.01