Frozen ViT embeddings in chest radiography suppress small-lesion signal at the CLS token but recover it via patch-local pooling on the same forward pass across multiple models and large cohorts.
Revisiting [cls] and patch token interaction in vision transformers
4 Pith papers cite this work. Polarity classification is still indexing.
fields
cs.CV 4years
2026 4verdicts
UNVERDICTED 4representative citing papers
Decouples semantic and spatial tokens in NVS transformers to resolve representation ambiguity, yielding consistent gains with near-zero added latency.
GeoCFNet uses a DINOv3 backbone with Token-Differentiated Fusion and Geometry-Aware Spatial Regularization to estimate confidence fields for ESD guidance, reporting RMSE 0.0480, PSNR 27.1995, SSIM 0.3397, and CC 0.2466.
Register tokens improve pixel-space Diffusion Transformers by cleaning high-noise feature maps, and Register Guidance amplifies that effect.
citing papers explorer
-
Frozen Foundation-Model Embeddings Discard Small-Lesion Signal in Chest Radiography: Implications for Pre-Deployment Evaluation
Frozen ViT embeddings in chest radiography suppress small-lesion signal at the CLS token but recover it via patch-local pooling on the same forward pass across multiple models and large cohorts.
-
Resolving Representation Ambiguity in Feedforward Novel View Synthesis Transformer via Semantic-Spatial Decoupling
Decouples semantic and spatial tokens in NVS transformers to resolve representation ambiguity, yielding consistent gains with near-zero added latency.
-
GeoCFNet: Geometry-Aware Confidence Field Network for Robot-Assisted Endoscopic Submucosal Dissection
GeoCFNet uses a DINOv3 backbone with Token-Differentiated Fusion and Geometry-Aware Spatial Regularization to estimate confidence fields for ESD guidance, reporting RMSE 0.0480, PSNR 27.1995, SSIM 0.3397, and CC 0.2466.
-
Registers Matter for Pixel-Space Diffusion Transformers
Register tokens improve pixel-space Diffusion Transformers by cleaning high-noise feature maps, and Register Guidance amplifies that effect.