Data to 14 Oct 2026

A Causal View on Robustness of Neural Networks

arXiv 2005.01095 ↗First posted 3 May 2020cs.LG stat.MLcited by 53

Not on the map. Why?

Abstract (arXiv)

We present a causal view on the robustness of neural networks against input manipulations, which applies not only to traditional classification tasks but also to general measurement data. Based on this view, we design a deep causal manipulation augmented model (deep CAMA) which explicitly models possible manipulations on certain causes leading to changes in the observed effect. We further develop data augmentation and test-time fine-tuning methods to improve deep CAMA's robustness. When compared with discriminative deep neural networks, our proposed model shows superior robustness against unseen manipulations. As a by-product, our model achieves disentangled representation which separates the representation of manipulations from those of other latent causes.

Key facts. Cited by 53 papers in the map (data through 14 Oct 2026). First posted on arXiv 3 May 2020.

Builds on

No reference list yet.

Built on by

Cited by year (2026 to 14 Oct)

Citing papers per year: 2021: 18; 2022: 11; 2023: 9; 2024: 7; 2025: 5; 2026: 3.

53 papers.

Sort:

By area:

2021Generalizing to Unseen Domains: A Survey on Domain Generalizationcited by 766
2023Backdoor Defense via Deconfounded Representation Learningcited by 38
2021CausalAdv: Adversarial Robustness through the Lens of Causalitycited by 37
2021Graph Domain Adaptation: A Generative Viewcited by 35
2024The Essential Role of Causality in Foundation World Models for Embodied AIcited by 32
2021Out-of-distribution Prediction with Invariant Risk Minimization: The Limitation and An Effective Fixcited by 27
2022Certified Robustness Against Natural Language Attacks by Causal Interventioncited by 24
2022Explicit Tradeoffs between Adversarial and Natural Distributional Robustnesscited by 20
2022Learning Attention Propagation for Compositional Zero-Shot Learningcited by 20
2021Embracing the Disharmony in Medical Imaging: A Simple and Effective Framework for Domain Adaptationcited by 19
2021VACA: Design of Variational Graph Autoencoders for Interventional and Counterfactual Queriescited by 19
2021Domain-Specific Bias Filtering for Single Labeled Domain Generalizationcited by 18

38 more

2021Interpreting the Robustness of Neural NLP Models to Textual Perturbationscited by 17
2023RS2G: Data-Driven Scene-Graph Extraction and Embedding for Robust Autonomous Perception and Scenario Understandingcited by 17
2022CaRTS: Causality-driven Robot Tool Segmentation from Vision and Kinematics Datacited by 16
2022Adversarial Robustness for Tabular Data through Cost and Utility Awarenesscited by 15
2023A Review of the Role of Causality in Developing Trustworthy AI Systemscited by 14
2021Adversarial Visual Robustness by Causal Interventioncited by 13
2021Enhancing Model Robustness and Fairness with Causality: A Regularization Approachcited by 11
2022Deep Causal Learning: Representation, Discovery and Inferencecited by 11
2023Emerging Synergies in Causality and Deep Generative Models: A Surveycited by 10
2023Enhancing Adversarial Contrastive Learning via Adversarial Invariant Regularizationcited by 10
2021Training a Resilient Q-Network against Observational Interferencecited by 9
2022A.I. Robustness: a Human-Centered Perspective on Technological Challenges and Opportunitiescited by 9
2021Disentanglement and Generalization Under Correlation Shiftscited by 8
2022On the Generalization and Adaption Performance of Causal Modelscited by 8
2023Moderately Distributional Exploration for Domain Generalizationcited by 8
2021Instrumental Variable-Driven Domain Generalization with Unobserved Confounderscited by 7
2021Pulling Up by the Causal Bootstraps: Causal Data Augmentation for Pre-training Debiasingcited by 6
2022Causal Discovery and Knowledge Injection for Contestable Neural Networks (with Appendices)cited by 6
2021Adversarial for Good? How the Adversarial ML Community's Values Impede Socially Beneficial Uses of Attackscited by 5
2022Rethinking Causality-driven Robot Tool Segmentation with Temporal Constraintscited by 5
2024The Gaussian Discriminant Variational Autoencoder (GdVAE): A Self-Explainable Model with Counterfactual Explanationscited by 5
2024Generalizable Single-Source Cross-modality Medical Image Segmentation via Invariant Causal Mechanismscited by 5
2025Causal-SAM-LLM: Large Language Models as Causal Reasoners for Robust Medical Segmentationcited by 5
2024Conditional Generative Models are Sufficient to Sample from Any Causal Effect Estimandcited by 4
2024CausalDiff: Causality-Inspired Disentanglement via Diffusion Model for Adversarial Defensecited by 4
2021Causally Constrained Data Synthesis for Private Data Releasecited by 3
2022Causal Information Bottleneck Boosts Adversarial Robustness of Deep Neural Networkcited by 3
2023Where and How to Attack? A Causality-Inspired Recipe for Generating Counterfactual Adversarial Examplescited by 3
2021Counterfactual Adversarial Learning with Representation Interpolationcited by 2
2025Language-Conditioned Representations and Mixture-of-Experts Policy for Robust Multi-Task Robotic Manipulationcited by 2
2025Counterfactual Visual Explanation via Causally-Guided Adversarial Steeringcited by 1
2026Knowledge-Guided Adversarial Training for Infrared Object Detection via Thermal Radiation Modelingcited by 1
2026Outsmarting the Chameleon: Counterfactual Decoupling for Tactical OOD Shifts in Live Streaming Risk Assessmentcited by 1
2021Automated Domain Discovery from Multiple Sources to Improve Zero-Shot Generalizationcited by 0
2023Rethinking Robust Contrastive Learning from the Adversarial Perspectivecited by 0
2023Measuring the Effect of Causal Disentanglement on the Adversarial Robustness of Neural Network Modelscited by 0
2024Certified Causal Defense with Generalizable Robustnesscited by 0
2024Causal Interpretability for Adversarial Robustness: A Hybrid Generative Classification Approachcited by 0

50 most cited of 53 All 53 (JSON)