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

Confidence Calibration on ReClor (test)

4.4ECE

ORCE

2.64414.49726.3538.203May 12, 2026
Updated 21d ago

Evaluation Results

MethodLinks
2026.05
4.40.6268.0510.0640.04179.4
2026.05
4.60.1332.7970.1610.13980
2026.05
4.70.3516.0470.0910.06979.4
2026.05
6.10.3928.9050.3060.16151.2
2026.05
6.20.5379.4310.0830.0679.4
2026.05
6.50.5448.0610.1690.0860.8
2026.05
7.30.3442.3680.3290.18451.2
2026.05
8.40.4468.3780.2010.11260.8
2026.05
12.40.0840.0620.1830.16179.6
2026.05
140.3433.1960.250.16160.8
2026.05
14.10.1911.7120.4080.26351.2
2026.05
16.40.0590.1870.190.16779.4
2026.05
16.70.3111.2060.1470.12680.2
2026.05
17.80.0530.2390.3730.28360.8
2026.05
270.0610.1770.4410.29851.6
2026.05
300.1240.0060.3890.27155.6
2026.05
33.80.0670.1340.4410.30452.4
2026.05
35.70.1536.2160.4380.29251
2026.05
37.60.0720.110.3550.26961.6
2026.05
46.80.0490.2750.4770.33251.2
2026.05
48.30.0290.5230.4890.34651.4