← 研踪 PaperTrailarXiv:2603.28113 · v3
The multiplicative Lipschitz bound as an inductive bias
首次提交 2026-03-30 15:20
版本更新 2026-11-18 14:36
雷达观测 2026-11-19 19:08
摘要
We study the problem of training a neural network to minimize a loss function given a Lipschitz constraint. While the most obvious way to estimate the Lipschitz constant of a neural network is to multiply the Lipschitz constants of its layers, this bound, which we term the _trivial bound_, is often considered impractically loose. We show that, on the contrary, it can be more effective than state-of-the-art methods (Sandwich, Almost-Orthogonal Layers, Convex Potential Layers, and SDP-based Lipschitz Layers) for Lipschitz-constrained learning. Our method combines the trivial bound with a novel activation function (the quadrature nonlinearity) to achieve lower cross-entropy over most of the tested classification bounds. For autoencoders constrained to be nonexpansive in a learned norm, it achieves the lowest pixel reconstruction error at every tested bottleneck width, improving on PCA by 12--24%. Our architecture, consisting of concatenated cosine and sine activations, is supported by a semi-rigorous theoretical analysis of Lipschitz bound looseness that may prove valuable in its own right.
观测到的版本
相邻记录可能跳过中间版本;首次发现 v2 及以上也会标为版本更新。
v3 · 2026-11-18 14:36 · 最新观测版本
The multiplicative Lipschitz bound as an inductive bias
We study the problem of training a neural network to minimize a loss function given a Lipschitz constraint. While the most obvious way to estimate the Lipschitz constant of a neural network is to multiply the Lipschitz constants of its layers, this bound, which we term the _trivial bound_, is often considered impractically loose. We show that, on the contrary, it can be more effective than state-of-the-art methods (Sandwich, Almost-Orthogonal Layers, Convex Potential Layers, and SDP-based Lipschitz Layers) for Lipschitz-constrained learning. Our method combines the trivial bound with a novel activation function (the quadrature nonlinearity) to achieve lower cross-entropy over most of the tested classification bounds. For autoencoders constrained to be nonexpansive in a learned norm, it achieves the lowest pixel reconstruction error at every tested bottleneck width, improving on PCA by 12--24%. Our architecture, consisting of concatenated cosine and sine activations, is supported by a semi-rigorous theoretical analysis of Lipschitz bound looseness that may prove valuable in its own right.
查看此版本解读 →查看此版本原文 ↗