On the extended Concept Binding Benchmark, all tested VLMs, including the generative Diffusion Classifier, fail to distinguish left/right relations in generalised zero-shot settings, suggesting models rely on object recognition rather than relational composition.
Title resolution pending
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.CV 1years
2025 1verdicts
CONDITIONAL 1representative citing papers
citing papers explorer
-
Evaluating Compositional Generalisation in VLMs and Diffusion Models
On the extended Concept Binding Benchmark, all tested VLMs, including the generative Diffusion Classifier, fail to distinguish left/right relations in generalised zero-shot settings, suggesting models rely on object recognition rather than relational composition.