SEIS: Subspace-based Equivariance and Invariance Scores for Neural Representations

ArXi:2602.04054v2 Announce Type: replace Understanding how neural representations respond to geometric transformations is essential for evaluating whether learned features preserve meaningful spatial structure. Existing approaches primarily assess robustness primarily by comparing model outputs under transformed inputs, offering limited insight into how geometric information is organized within internal representations and failing to distinguish between information loss and re-encoding. In this work, we.