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

Visual Question Answering on HallusionBench

75.2Simple Accuracy

SpatialThinker-30B

45.14452.94760.7568.553Nov 10, 2025Nov 21, 2025Dec 3, 2025Dec 14, 2025Dec 26, 2025Jan 6, 2026Jan 18, 2026
Updated 18d ago

Evaluation Results

MethodLinks
2025.11
75.2--
2025.11
73.8--
2025.11
71.2--
2025.11
68.9--
2025.11
66.4--
2025.11
66.2--
2025.11
65.4--
2025.11
62.9--
2025.11
62.5--
2025.11
61.5--
2025.11
60.7--
2025.11
59--
2025.11
58.9--
2025.11
55.5--
2025.11
55--
2025.11
52.9--
2026.01
51.3115.5865.62
2026.01
50.8915.8167.93
2025.11
46.3--