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

Multi-image Spatial Reasoning on MindCube-Tiny (test)

69.2Aro Accuracy

GPT-5.2

19.48832.39445.358.206Feb 5, 2026Feb 6, 2026Feb 7, 2026Feb 8, 2026Feb 9, 2026
Updated 4d ago

Evaluation Results

MethodLinks
2026.02
69.295.541.558.4
2026.02
65.63150.250.2
2026.02
63.27531.6747.43
2026.02
59.649.547.250.6
2026.02
57.633.545.245.9
2026.02
56.431.547.846.8
2026.02
522934.537.6
2026.02
51.654.54649
2026.02
5033.539.240.7
2026.02
46.434.535.337.8
2026.02
44.84241.242.2
2026.02
443038.538.2
2026.02
43.641.538.240
2026.02
3833.532.734.1
2026.02
31.631.528.729.9
2026.02
25.63720.324.8
2026.02
21.438.729.529.3