Computer vision robustness
The Shape of Noise: Layer-Wise Perturbation Profiles for Diagnosing Vision Robustness
Question
How do perturbations change internal representations across a vision model's layers, and can those layer-wise profiles help diagnose where robustness breaks?
Public status
The project is ongoing. I am not listing a publication, submission status, contribution breakdown, paper, or code link because no public source currently verifies those details.
Current boundary
This page will expand when methods, results, authorship, and an external artifact can be cited publicly. Private drafts and review material are intentionally excluded.