Our new X account is live! Follow @wizwand_team for updates
WorkDL logo mark

Hallucination Detection on FELM worldknowledge (test)

88.3Nonfact Accuracy

Doc

-0.72422.38845.568.612Feb 25, 2024
Updated 4d ago

Evaluation Results

MethodLinks
2024.02
88.340.864.6
2024.02
81.617.949.8
2024.02
72.83453.4
2024.02
70.729.950.3
2024.02
54.485.569.9
2024.02
42.99468.4
2024.02
40.862.351.6
2024.02
40.179.259.7
2024.02
35.493.264.3
2024.02
34.773.254
2024.02
21.895.358.5
2024.02
11.694.352.9
2024.02
8.895.152
2024.02
3.496.149.8
2024.02
2.796.949.8