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

Question Answering Calibration on NQ, TQ, HQ, 2Wiki, Pararel

0.05Calibration Error (NQ)

EliCal(1k)

0.03560.13280.230.3272Oct 20, 2025
Updated 3mo ago

Evaluation Results

MethodLinks
2025.10
0.050.060.030.050.080.05
2025.10
0.060.050.040.050.040.06
2025.10
0.070.070.160.170.030.12
2025.10
0.090.050.10.140.10.1
2025.10
0.090.030.060.240.210.13
2025.10
0.120.160.180.130.130.15
2025.10
0.130.040.110.280.250.16
2025.10
0.270.230.260.150.30.22
2025.10
0.390.530.320.310.460.4
2025.10
0.390.240.470.50.320.39
2025.10
0.410.280.440.410.280.37