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

Spatial Reasoning on SPBench MV

68.2NQ Score

SpatialLadder-3B

18.69631.54844.457.252May 19, 2026
Updated 14d ago

Evaluation Results

MethodLinks
2026.05
68.27570.9
2026.05
63.758.953.1
2026.05
63.253.759.4
2026.05
61.272.865.8
2026.05
56.259.657.6
2026.05
51.950.756.5
2026.05
40.759.448.2
2026.05
37.751.443.2
2026.05
37.347.541.4
2026.05
32.55340.7
2026.05
28.949.837.3
2026.05
25.653.236.6
2026.05
23.357.637
2026.05
20.649.632.2