Active Spatial Guidance replaces injected positional embeddings in ViTs with a training-only 2D coordinate regression loss on final-layer tokens, yielding better results than learned absolute or rotary embeddings on ImageNet-100, ADE20K, and Hypersim under matched training.
Title resolution pending
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.CV 1years
2026 1verdicts
UNVERDICTED 1representative citing papers
citing papers explorer
-
Active Spatial Guidance: Eliminating Injected Positional Mechanisms in Vision Transformers
Active Spatial Guidance replaces injected positional embeddings in ViTs with a training-only 2D coordinate regression loss on final-layer tokens, yielding better results than learned absolute or rotary embeddings on ImageNet-100, ADE20K, and Hypersim under matched training.