Binary Spiking Neural Networks as Causal Models
Not on the map. Why?
Abstract (arXiv)
We provide a causal analysis of Binary Spiking Neural Networks (BSNNs) to explain their behavior. We formally define a BSNN and represent its spiking activity as a binary causal model. Thanks to this causal representation, we are able to explain the output of the network by leveraging logic-based methods. In particular, we show that we can successfully use a SAT as well as a SMT solver to compute abductive explanations from this binary causal model. To illustrate our approach, we trained the BSNN on the standard MNIST dataset and applied our SAT-based and SMT-based methods to finding abductive explanations of the network's classifications based on pixel-level features. We also compared the found explanations against SHAP, a popular method used in the area of explainable AI. We show that, unlike SHAP, our approach guarantees that a found explanation does not contain completely irrelevant features.
Key facts. Cited by 1 paper in the map (data through 7 Oct 2026). Builds on 26 of its 44 references that are in the map. First posted on arXiv 29 Apr 2026.
Builds on
26 of 44 references.
References · 26
- 2017A Unified Approach to Interpreting Model Predictionscited by 6,488
- 2013Estimating or Propagating Gradients Through Stochastic Neurons for Conditional Computationcited by 2,031
- 2016XNOR-Net: ImageNet Classification Using Binary Convolutional Neural Networkscited by 829
- 2019Surrogate Gradient Learning in Spiking Neural Networkscited by 604
- 2020Binary Neural Networks: A Surveycited by 184
- 2018Abduction-Based Explanations for Machine Learning Modelscited by 159
- 2018A Symbolic Approach to Explaining Bayesian Network Classifierscited by 130
- 2020On The Reasons Behind Decisionscited by 99
- 2018Contrastive Explanation: A Structural-Model Approachcited by 64
- 2021On Explaining Random Forests with SATcited by 57
- 2020Explaining Naive Bayes and Other Linear Classifiers with Polynomial Time and Delaycited by 56
- 2023SpikingBERT: Distilling BERT to Train Spiking Language Models Using Implicit Differentiationcited by 47
14 more
- 2020Interpreting Neural Networks as Gradual Argumentation Frameworks (Including Proof Appendix)cited by 38
- 2020On Tractable Representations of Binary Neural Networkscited by 34
- 2020Exploring the Connection Between Binary and Spiking Neural Networkscited by 33
- 2020BS4NN: Binarized Spiking Neural Networks with Temporal Coding and Learningcited by 20
- 2023SpArX: Sparse Argumentative Explanations for Neural Networks [Technical Report]cited by 18
- 2021A unified logical framework for explanations in classifier systemscited by 15
- 2014The Computational Complexity of Structure-Based Causalitycited by 12
- 2019ReStoCNet: Residual Stochastic Binary Convolutional Spiking Neural Network for Memory-Efficient Neuromorphic Computingcited by 12
- 2020The Counterfactual NESS Definition of Causationcited by 8
- 2024From SHAP Scores to Feature Importance Scorescited by 8
- 2020BiSNN: Training Spiking Neural Networks with Binary Weights via Bayesian Learningcited by 6
- 2023A Refutation of Shapley Values for Explainabilitycited by 4
- 2023Refutation of Shapley Values for XAI -- Additional Evidencecited by 4
- 2024Relevant Irrelevance: Generating Alterfactual Explanations for Image Classifierscited by 3
Built on by
Citing papers per year: 2026: 1.
1 paper.