Sparse autoencoders trained on Mammo-CLIP features expose a small set of concept-aligned and confounding latent neurons in breast cancer predictions.
Title resolution pending
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.CV 1years
2025 1verdicts
CONDITIONAL 1representative citing papers
citing papers explorer
-
Mammo-SAE: Interpreting Breast Cancer Concept Learning with Sparse Autoencoders
Sparse autoencoders trained on Mammo-CLIP features expose a small set of concept-aligned and confounding latent neurons in breast cancer predictions.