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

Multi-image Hallucination Evaluation on BLINK

61.33Accuracy

GLM4.1VBase + CAPL

40.384445.822251.2656.6978Mar 7, 2026
Updated 1mo ago

Evaluation Results

MethodLinks
2026.03
61.33
2026.03
58.17
2026.03
57.76
2026.03
55.76
2026.03
54.81
2026.03
54.6
2026.03
53.17
2026.03
50.34
2026.03
48.34
2026.03
45.24
2026.03
44.77
2026.03
41.19