# Causal Learning and Explanation of Deep Neural Networks via Autoencoded Activations

arXiv: https://arxiv.org/abs/1802.00541 · HTML: https://mlsanity.com/paper/1802.00541/

## Key facts

- Cited by 23 papers in the map (data through 14 Oct 2026).
- First posted on arXiv 2 Feb 2018.

## Built on by (most cited first, 23 of 23)

- [Causal Abstraction: A Theoretical Foundation for Mechanistic Interpretability](https://mlsanity.com/paper/2301.04709/) (2023)
- [Recent Advancements in End-to-End Autonomous Driving using Deep Learning: A Survey](https://mlsanity.com/paper/2307.04370/) (2023)
- [Explainability of deep vision-based autonomous driving systems: Review and challenges](https://mlsanity.com/paper/2101.05307/) (2021)
- [Causality-based Neural Network Repair](https://mlsanity.com/paper/2204.09274/) (2022)
- [Physics-Guided Deep Learning for Dynamical Systems: A Survey](https://mlsanity.com/paper/2107.01272/) (2021)
- [STEEX: Steering Counterfactual Explanations with Semantics](https://mlsanity.com/paper/2111.09094/) (2021)
- [Defining and Quantifying the Emergence of Sparse Concepts in DNNs](https://mlsanity.com/paper/2111.06206/) (2021)
- [Inspect, Understand, Overcome: A Survey of Practical Methods for AI Safety](https://mlsanity.com/paper/2104.14235/) (2021)
- [OCTET: Object-aware Counterfactual Explanations](https://mlsanity.com/paper/2211.12380/) (2022)
- [Causal Reasoning Meets Visual Representation Learning: A Prospective Study](https://mlsanity.com/paper/2204.12037/) (2022)
- [Evaluation Methods and Measures for Causal Learning Algorithms](https://mlsanity.com/paper/2202.02896/) (2022)
- [Towards Compositional Interpretability for XAI](https://mlsanity.com/paper/2406.17583/) (2024)
- [Can We Faithfully Represent Masked States to Compute Shapley Values on a DNN?](https://mlsanity.com/paper/2105.10719/) (2021)
- [Detecting Multi-Sensor Fusion Errors in Advanced Driver-Assistance Systems](https://mlsanity.com/paper/2109.06404/) (2021)
- [Responsible and Regulatory Conform Machine Learning for Medicine: A Survey of Challenges and Solutions](https://mlsanity.com/paper/2107.09546/) (2021)
- [Explaining Language Models' Predictions with High-Impact Concepts](https://mlsanity.com/paper/2305.02160/) (2023)
- [SUNY: A Visual Interpretation Framework for Convolutional Neural Networks from a Necessary and Sufficient Perspective](https://mlsanity.com/paper/2303.00244/) (2023)
- [GIFT: A Framework Towards Global Interpretable Faithful Textual Explanations of Vision Classifiers](https://mlsanity.com/paper/2411.15605/) (2024)
- [Explainability-aided Domain Generalization for Image Classification](https://mlsanity.com/paper/2104.01742/) (2021)
- [Explanatory causal effects for model agnostic explanations](https://mlsanity.com/paper/2206.11529/) (2022)
- [Neural Bayesian Network Understudy](https://mlsanity.com/paper/2211.08243/) (2022)
- [Causally-informed Deep Learning towards Explainable and Generalizable Outcomes Prediction in Critical Care](https://mlsanity.com/paper/2502.02109/) (2025)
- [A Model of Causal Explanation on Neural Networks for Tabular Data](https://mlsanity.com/paper/2512.21746/) (2025)

## Abstract (arXiv)

Deep neural networks are complex and opaque. As they enter application in a variety of important and safety critical domains, users seek methods to explain their output predictions. We develop an approach to explaining deep neural networks by constructing causal models on salient concepts contained in a CNN. We develop methods to extract salient concepts throughout a target network by using autoencoders trained to extract human-understandable representations of network activations. We then build a bayesian causal model using these extracted concepts as variables in order to explain image classification. Finally, we use this causal model to identify and visualize features with significant causal influence on final classification.