Figureure 1 · Sparse autoencoders approximate nonlinear manifolds (dark blue, mostlyFigure 1. Sparse autoencoders approximate nonlinear manifolds (dark blue, mostly occluded) with linear patches (light blue). We show that identifiability hinges on four key ingredients: (a) the approximation being good enough (low reconstruction error), (b) the manifold being sampled densely enough, (c) co-occurring concepts being distinct enough (an approximate restricted isometry property), and (d) sufficiently diverse concept co-occurrence patterns. When these hold, individual patches are identifiable, rendering the whole model statistically identifiable.这张图来自论文 PDF 的结构化抽取。当前用于辅助理解 Toward Identifiable Sparse Autoencoders 的方法或实验,请结合正文精读段落一起看。