mlsanityData to 14 Oct 2026

Deterministic versus stochastic dynamical classifiers: opposing random adversarial attacks with noise

arXiv 2409.13470 ↗First posted 20 Sep 2024cs.LG cond-mat.dis-nn cond-mat.stat-mechcited by 4

Not on the map. Why?

Abstract (arXiv)

The Continuous-Variable Firing Rate (CVFR) model, widely used in neuroscience to describe the intertangled dynamics of excitatory biological neurons, is here trained and tested as a veritable dynamically assisted classifier. To this end the model is supplied with a set of planted attractors which are self-consistently embedded in the inter-nodes coupling matrix, via its spectral decomposition. Learning to classify amounts to sculp the basin of attraction of the imposed equilibria, directing different items towards the corresponding destination target, which reflects the class of respective pertinence. A stochastic variant of the CVFR model is also studied and found to be robust to aversarial random attacks, which corrupt the items to be classified. This remarkable finding is one of the very many surprising effects which arise when noise and dynamical attributes are made to mutually resonate.

Key facts. Cited by 4 papers in the map (data through 14 Oct 2026). Builds on 17 of its 42 references that are in the map. First posted on arXiv 20 Sep 2024.

Builds on

17 of 42 references.

References · 17

  1. 2018Neural Ordinary Differential Equationscited by 3,779
  2. 2020Hopfield Networks is All You Needcited by 475
  3. 2020Time Series Forecasting With Deep Learning: A Surveycited by 459
  4. 2020Machine Learning Force Fieldscited by 207
  5. 2019Recurrent Neural Networks for Time Series Forecasting: Current Status and Future Directionscited by 172
  6. 2020Liquid Time-constant Networkscited by 161
  7. 2019Neural SDE: Stabilizing Neural ODE Networks with Stochastic Noisecited by 101
  8. 2015Learning Natural Language Inference with LSTMcited by 51
  9. 2017Dense Associative Memory is Robust to Adversarial Inputscited by 45
  10. 2020Machine learning in spectral domaincited by 20
  11. 2020Solving non-linear Kolmogorov equations in large dimensions by using deep learning: a numerical comparison of discretization schemescited by 11
  12. 2019Defense Methods Against Adversarial Examples for Recurrent Neural Networkscited by 8
5 more
  1. 2023How a student becomes a teacher: learning and forgetting through Spectral methodscited by 5
  2. 2023Stable Attractors for Neural networks classification via Ordinary Differential Equations (SA-nODE)cited by 5
  3. 2023Complex Recurrent Spectral Networkcited by 5
  4. 2023Training neural networks with structured noise improves classification and generalizationcited by 4
  5. 2024Estimating Global Input Relevance and Enforcing Sparse Representations with a Scalable Spectral Neural Network Approachcited by 4

Built on by

Cited by year (2026 to 14 Oct)

Citing papers per year: 2026: 4.

4 papers.

  1. 2026Exact Fixed-Point Constraints in Neural-ODEs with Provable Universalitycited by 1
  2. 2026Approximating velocity fields with planted attractors via Neural-ODEs for classification purposescited by 0
  3. 2026Spectral Higher-Order Neural Networks Have Sharp Expressivity Boundscited by 0
  4. 2026Multiplicity of Stable Attractors in Disordered Neural Modelscited by 0