2 papers
cs.CV2026
How Vision Transformers Represent Position: A Mechanistic Comparison of Absolute and Rotary Positional Encodings
Mahmoud Mannes, Nikhil Maturi, Parva Mehta +2
Positional embeddings encode the spatial position of each image patch in a Vision Transformer (ViT), allowing the model to distinguish different spatial arrangements of visual cont…
cs.CV2026
Positional Encodings Anchor Spatial Structure in Vision Transformers: A Geometric Perspective on Robustness
Mahmoud Mannes
Positional embeddings (PEs) in Vision Transformers (ViTs) are known to impact performance and robustness, but their role in shaping internal spatial representations is not well und…