Pith. sign in

REVIEW 3 cited by

Some Fundamental Aspects about Lipschitz Continuity of Neural Networks

Not yet reviewed by Pith; the record is open.

This paper has not been read by Pith yet. Machine review is queued; the pith claim, tier, and objections will appear here once it completes.

SPECIMEN: schema-true, not a live event

T0 review · schema-true

One-sentence machine reading of the paper's core claim.

pith:XXXXXXXX · record.json · timestamp

arxiv 2302.10886 v4 pith:DWY4UMBR submitted 2023-02-21 cs.LG stat.ML

Some Fundamental Aspects about Lipschitz Continuity of Neural Networks

classification cs.LG stat.ML
keywords lipschitzboundslowercontinuitydifferentgeneralisationinvestigationlabel
verification ladder T0 review T1 audit T2 compute T3 formal T4 reserved
0 comments
Share X Bluesky LinkedIn Reddit HN
read the original abstract

Lipschitz continuity is a crucial functional property of any predictive model, that naturally governs its robustness, generalisation, as well as adversarial vulnerability. Contrary to other works that focus on obtaining tighter bounds and developing different practical strategies to enforce certain Lipschitz properties, we aim to thoroughly examine and characterise the Lipschitz behaviour of Neural Networks. Thus, we carry out an empirical investigation in a range of different settings (namely, architectures, datasets, label noise, and more) by exhausting the limits of the simplest and the most general lower and upper bounds. As a highlight of this investigation, we showcase a remarkable fidelity of the lower Lipschitz bound, identify a striking Double Descent trend in both upper and lower bounds to the Lipschitz and explain the intriguing effects of label noise on function smoothness and generalisation.

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.

Forward citations

Cited by 3 Pith papers

Reviewed papers in the Pith corpus that reference this work. Sorted by Pith novelty score.

  1. Trainable Nonexpansive Denoisers for Contractive Image Reconstruction

    eess.IV 2026-07 conditional novelty 7.0

    Permutation-symmetrized positive-weight aggregation creates trainable denoisers that are globally nonexpansive and make plug-and-play reconstruction a contraction.

  2. A guided residual search for nonlinear state-space identification

    eess.SP 2026-02 conditional novelty 6.0

    A three-stage identification method (guided residual search, neural-network residual learning, multiple-shooting refinement) makes nonlinear state-space model fitting more reliable.

  3. Scalable Equilibrium Propagation via Intermediate Error Signals for Deep Convolutional CRNNs

    cs.LG 2025-08 unverdicted novelty 6.0

    Introduces layer-wise learning signals combining knowledge distillation and local errors into Equilibrium Propagation, enabling scalable training of deep VGG-style CRNNs with SOTA results on CIFAR-10 and CIFAR-100.