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

Abstract Visual Reasoning on ARC-AGI 1

98Accuracy (Pass@2)

best.human

12.51234.70656.979.094Feb 2, 2026Feb 13, 2026Feb 24, 2026Mar 7, 2026Mar 18, 2026Mar 29, 2026Apr 9, 2026
Updated 9d ago

Evaluation Results

MethodLinks
2026.02
98
2026.02
79.6
2026.04
79.6
2026.02
66.7
2026.04
66.7
2026.02
65.8
2026.02
63.8
2026.02
60.4
2026.02
60.2
2026.02
60.1
2026.04
56.8
2026.04
54.7
2026.02
54.5
2026.04
54.5
2026.04
53
2026.02
44.6
2026.04
44.6
2026.02
44
2026.04
44
2026.02
40.3
2026.04
40.3
34.5
2026.04
34.5
21.2
2026.04
21.2
2026.02
15.8
2026.04
15.8