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

Compositional Reasoning on Lego Puzzles

57.7Accuracy

GPT-4o-0513

25.77234.06142.3550.639Nov 10, 2025
Updated 18d ago

Evaluation Results

MethodLinks
57.7
53.6
2025.11
37.7
2025.11
36.6
35.8
2025.11
34.7
2025.11
33.9
33.4
2025.11
32
31.5
29.9
2025.11
29.7
2025.11
27