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

Long-form Generation Hallucination Evaluation on CHAIR

58.8CS Score

VCD

10.33622.91835.548.082May 23, 2026
Updated 8d ago

Evaluation Results

MethodLinks
2026.05
58.817.570.1
2026.05
54.617.371.3
2026.05
52.414.573.8
2026.05
51.613.875
2026.05
5013.176.2
2026.05
49.413.176.4
2026.05
43.414.377.2
2026.05
42.614.274.7
2026.05
41.415.673.8
2026.05
3915.876.5
2026.05
32.311.675.7
2026.05
29.29.784.5
2026.05
28.214.779.4
2026.05
27.413.275.5
2026.05
27.28.582.5
2026.05
26.67.486.2
2026.05
25.26.970.1
2026.05
257.787.1
2026.05
247.387.4
2026.05
23.27.585.5
2026.05
22.46.488.9
2026.05
21.811.784.1
2026.05
21.64.989.3
2026.05
20.810.785.8
2026.05
20.27.288.9
2026.05
19.610.686
2026.05
19.25.990.4
2026.05
19.29.883.6
2026.05
18.65.390.5
2026.05
16.47.490.2
2026.05
15.67.289.7
2026.05
15.46.790.6
2026.05
153.692
2026.05
14.46.690
2026.05
13492.6
2026.05
12.2691.7