cs.CRNov 13, 2026

Binarization as an Attack Surface in Weightless Neural Networks

Authors: Lucas Senos Coutinho, Matheus Vargas, Pedro Teixeira, Heudson Tosta Mirandola, Ori Rottenstreich, Priscila Machado Vieira Lima, Leandro Santiago, Daniel Sadoc Menasche

Organizations: Universidade Federal do Rio de Janeiro (UFRJ), Rio de Janeiro, Brazil · Elytron Cybersecurity · Technion–Israel Institute of Technology, Haifa, Israel · Universidade Federal Fluminense (UFF), Niter´oi, Brazil

Abstract

Weightless Neural Networks (WNNs) are lightweight classifiers that binarize their inputs and learn by storing binary patterns in lookup memories, which makes them attractive for resource-constrained tasks such as intrusion detection. Because a small perturbation that crosses no binarization threshold leaves the binary input unchanged, WNNs may appear naturally robust to adversarial examples; a perturbation that does cross a threshold, however, can still change the prediction. We study this tension in WiSARD, the canonical RAM-based WNN, and in two variants that replace its memories with Bloom filters, Bloom WiSARD and BTHOWeN. A continuous perturbation reaches these classifiers only by crossing an encoding threshold, changing at least one binary tuple, and redirecting a memory query; we formalize this mechanism with a threshold-crossing bound and a response-margin certificate for standard Boolean WiSARD. Controlled MNIST experiments compare linear and data-driven thermometer encodings under model-agnostic sensitivity probes, a score-based Simulated Annealing attack, and surrogate-transfer attacks. Changing only the encoding while keeping each model otherwise fixed shows that encoding granularity can trade a small clean-accuracy loss for a large decrease in attack success. On the Bot-IoT, CICIDS, and UNSW-NB15 intrusion-detection datasets, adversarial perturbations lead WNN detectors to classify many attacks as normal traffic and to raise false alarms, even when their aggregate accuracy remains high.

Figures & tables

Explore similar work

CardsList
  1. A Few Large Shifts: Layer-Inconsistency Based Minimal Overhead Adversarial Example Detection

    May 19, 2025Sanggeon Yun, Ryozo Masukawa, Hyunwoo Oh +2Adversarial AttacksAdversarial Examples

  2. Enhancing Adversarial Robustness in Network Intrusion Detection: A Layer-wise Adaptive Regularization Approach

    May 9, 2026Hira Nasir, Eiman Javed, Balawal Shabir +2Adversarial TrainingNetwork Intrusion Detection

  3. Weight Oracles: Reading Neural Network Weights with Language Models

    Oct 5, 2026Krishna Kabra, Constantin Venhoff, Christian Schroeder de WittLLM AuditingBackdoor Detection