IQBench evaluates seven vision-language models on 500 visual IQ questions and finds they struggle most with 3D spatial and anagram reasoning.
Silvar-med: A speech-driven visual language model for explainable abnormality detection in medical imaging
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
citation-role summary
background 1
citation-polarity summary
fields
cs.CV 1years
2025 1verdicts
CONDITIONAL 1roles
background 1polarities
unclear 1representative citing papers
citing papers explorer
-
IQBench: How "Smart'' Are Vision-Language Models? A Study with Human IQ Tests
IQBench evaluates seven vision-language models on 500 visual IQ questions and finds they struggle most with 3D spatial and anagram reasoning.