The Geometry of Representational Failures in Vision Language Models

Documenti analoghi

• SAEmnesia: Erasing Concepts in Diffusion Models with Supervised Sparse Autoencoders
  di: Cassano, Enrico, et al.
  Pubblicazione: (2025)
• HOLMES: HOLonym-MEronym based Semantic inspection for Convolutional Image Classifiers
  di: Dibitonto, Francesco, et al.
  Pubblicazione: (2024)
• Binding Visual Features Point by Point
  di: Haputhanthri, Udith, et al.
  Pubblicazione: (2026)
• Attributes Shape the Embedding Space of Face Recognition Models
  di: Leroy, Pierrick, et al.
  Pubblicazione: (2025)
• Understanding the Limits of Vision Language Models Through the Lens of the Binding Problem
  di: Campbell, Declan, et al.
  Pubblicazione: (2024)