#imageclustering
New preprint! 🚀

How well do the internal representations of VLMs organize visual information for #unsupervised #imageclustering?

We explore this over 252 configurations, with some interesting results!

arxiv.org/abs/2610.33265

#VLM #visionlanguagemodels #INESCINOV #ULusófona #FCUL #LASIGE
Image Clustering with Vision-Language Model Embeddings
Pretrained vision-language models provide image representations whose suitability for clustering may vary with feature extraction, image domain, and partitioning algorithm. We compare nine representat...
arxiv.org
October 27, 2026 at 5:13 PM