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

Visual Reasoning on VisualPuzzles OOD (test)

47.95Overall Accuracy

Gemini-3.1Pro

30.63435.129539.62544.1205May 12, 2026
Updated 21d ago

Evaluation Results

MethodLinks
47.9560.6961.145228.2338.11
2026.05
41.349.258.34927.326.2
2026.05
35.0245.830.8147.526.7925.52
2026.05
34.4233.5929.864929.6731.82
2026.05
32.9637.7927.014126.7927.01
2026.05
32.8837.7929.3844.528.7125.87
2026.05
32.7137.0221.847.526.3221.8
2026.05
31.332.423.738.528.732.5