# Causal Discovery and Knowledge Injection for Contestable Neural Networks (with Appendices)

arXiv: https://arxiv.org/abs/2205.09787 · HTML: https://mlsanity.com/paper/2205.09787/

## Key facts

- Cited by 6 papers in the map (data through 14 Oct 2026).
- Builds on 17 of its 34 references that are in the map.
- First posted on arXiv 19 May 2022.

## Builds on (most cited first, 17 of 17)

- [Explaining and Harnessing Adversarial Examples](https://mlsanity.com/paper/1412.6572/) (2014)
- [Intriguing properties of neural networks](https://mlsanity.com/paper/1312.6199/) (2013)
- [DAGs with NO TEARS: Continuous Optimization for Structure Learning](https://mlsanity.com/paper/1803.01422/) (2018)
- [Inductive Biases for Deep Learning of Higher-Level Cognition](https://mlsanity.com/paper/2011.15091/) (2020)
- [A Survey of Human-in-the-loop for Machine Learning](https://mlsanity.com/paper/2108.00941/) (2021)
- [Informed Machine Learning -- A Taxonomy and Survey of Integrating Knowledge into Learning Systems](https://mlsanity.com/paper/1903.12394/) (2019)
- [Explainable Machine Learning for Scientific Insights and Discoveries](https://mlsanity.com/paper/1905.08883/) (2019)
- [Beware of the Simulated DAG! Causal Discovery Benchmarks May Be Easy To Game](https://mlsanity.com/paper/2102.13647/) (2021)
- [Inducing Causal Structure for Interpretable Neural Networks](https://mlsanity.com/paper/2112.00826/) (2021)
- [A Causal View on Robustness of Neural Networks](https://mlsanity.com/paper/2005.01095/) (2020)
- [Explanation-Based Human Debugging of NLP Models: A Survey](https://mlsanity.com/paper/2104.15135/) (2021)
- [CASTLE: Regularization via Auxiliary Causal Graph Discovery](https://mlsanity.com/paper/2009.13180/) (2020)
- [FIND: Human-in-the-Loop Debugging Deep Text Classifiers](https://mlsanity.com/paper/2010.04987/) (2020)
- [The impact of prior knowledge on causal structure learning](https://mlsanity.com/paper/2102.00473/) (2021)
- [Improving Deep Learning Models via Constraint-Based Domain Knowledge: a Brief Survey](https://mlsanity.com/paper/2005.10691/) (2020)
- [Combining Experts' Causal Judgments](https://mlsanity.com/paper/2005.10180/) (2020)
- [Contestable Black Boxes](https://mlsanity.com/paper/2006.05133/) (2020)

## Built on by (most cited first, 6 of 6)

- [Contestable AI needs Computational Argumentation](https://mlsanity.com/paper/2405.10729/) (2024)
- [CAUSAL3D: A Comprehensive Benchmark for Causal Learning from Visual Data](https://mlsanity.com/paper/2503.04852/) (2025)
- [Argumentative Causal Discovery](https://mlsanity.com/paper/2405.11250/) (2024)
- [Contestability in Quantitative Argumentation](https://mlsanity.com/paper/2507.11323/) (2025)
- [Online Handbook of Argumentation for AI: Volume 4](https://mlsanity.com/paper/2401.09444/) (2023)
- [Facets of Disparate Impact: Evaluating Legally Consistent Bias in Machine Learning](https://mlsanity.com/paper/2505.05471/) (2025)

## Abstract (arXiv)

Neural networks have proven to be effective at solving machine learning tasks but it is unclear whether they learn any relevant causal relationships, while their black-box nature makes it difficult for modellers to understand and debug them. We propose a novel method overcoming these issues by allowing a two-way interaction whereby neural-network-empowered machines can expose the underpinning learnt causal graphs and humans can contest the machines by modifying the causal graphs before re-injecting them into the machines. The learnt models are guaranteed to conform to the graphs and adhere to expert knowledge, some of which can also be given up-front. By building a window into the model behaviour and enabling knowledge injection, our method allows practitioners to debug networks based on the causal structure discovered from the data and underpinning the predictions. Experiments with real and synthetic tabular data show that our method improves predictive performance up to 2.4x while producing parsimonious networks, up to 7x smaller in the input layer, compared to SOTA regularised networks.