AI RESEARCH
Beyond Compression: Quantifying Spectral Accessibility in Vision Representations
arXiv CS.CV
•
ArXi:2606.03795v1 Announce Type: new Vision-language models map visual features into a shared embedding space through learned projection layers, yet it remains unclear how these transformations alter the structure of visual information. This study examines changes in representation through spatial-frequency accessibility, measured by the linear recoverability of band-limited Fourier energy from model representations. To isolate effects beyond dimensionality reduction, we