Lens: Bringing the Right Semantic Perspective into Focus for Training-Free Multimodal Representation Learning.

arXiv:2609.20252v1 Announce Type: cross Abstract: High-quality representations are essential for a wide range of downstream tasks. Dedicated embedding models are explicitly optimized for representation learning, yet their training data are often more limited i