# Policy Optimization with Sparse Global Contrastive Explanations

arXiv: https://arxiv.org/abs/2207.06269 · HTML: https://mlsanity.com/paper/2207.06269/

## Key facts

- Cited by 2 papers in the map (data through 14 Oct 2026).
- Builds on 23 of its 33 references that are in the map.
- First posted on arXiv 13 Jul 2022.

## Builds on (most cited first, 23 of 23)

- [Towards A Rigorous Science of Interpretable Machine Learning](https://mlsanity.com/paper/1702.08608/) (2017)
- [Explanation in Artificial Intelligence: Insights from the Social Sciences](https://mlsanity.com/paper/1706.07269/) (2017)
- [Counterfactual Explanations without Opening the Black Box: Automated Decisions and the GDPR](https://mlsanity.com/paper/1711.00399/) (2017)
- [Dynamic Routing Between Capsules](https://mlsanity.com/paper/1710.09829/) (2017)
- [Interpretable Machine Learning: Fundamental Principles and 10 Grand Challenges](https://mlsanity.com/paper/2103.11251/) (2021)
- [Actionable Recourse in Linear Classification](https://mlsanity.com/paper/1809.06514/) (2018)
- [Data-Efficient Off-Policy Policy Evaluation for Reinforcement Learning](https://mlsanity.com/paper/1604.00923/) (2016)
- [Techniques for Interpretable Machine Learning](https://mlsanity.com/paper/1808.00033/) (2018)
- [Model-Agnostic Counterfactual Explanations for Consequential Decisions](https://mlsanity.com/paper/1905.11190/) (2019)
- [Programmatically Interpretable Reinforcement Learning](https://mlsanity.com/paper/1804.02477/) (2018)
- [Interpretable Convolutional Neural Networks](https://mlsanity.com/paper/1710.00935/) (2017)
- [Visualizing and Understanding Recurrent Networks](https://mlsanity.com/paper/1506.02078/) (2015)
- [Explainable Reinforcement Learning: A Survey](https://mlsanity.com/paper/2005.06247/) (2020)
- [Boolean Decision Rules via Column Generation](https://mlsanity.com/paper/1805.09901/) (2018)
- [Contrastive Explanations for Model Interpretability](https://mlsanity.com/paper/2103.01378/) (2021)
- [Toward Interpretable Deep Reinforcement Learning with Linear Model U-Trees](https://mlsanity.com/paper/1807.05887/) (2018)
- [Contrastive Explanations for Reinforcement Learning in terms of Expected Consequences](https://mlsanity.com/paper/1807.08706/) (2018)
- [Bootstrapping Fitted Q-Evaluation for Off-Policy Inference](https://mlsanity.com/paper/2102.03607/) (2021)
- [Contrastive Explanations with Local Foil Trees](https://mlsanity.com/paper/1806.07470/) (2018)
- [Generating Counterfactual and Contrastive Explanations using SHAP](https://mlsanity.com/paper/1906.09293/) (2019)
- [What Did You Think Would Happen? Explaining Agent Behaviour Through Intended Outcomes](https://mlsanity.com/paper/2011.05064/) (2020)
- [Optimization Methods for Interpretable Differentiable Decision Trees in Reinforcement Learning](https://mlsanity.com/paper/1903.09338/) (2019)
- [Tradeoff-Focused Contrastive Explanation for MDP Planning](https://mlsanity.com/paper/2004.12960/) (2020)

## Built on by (most cited first, 2 of 2)

- [Relative Sparsity for Medical Decision Problems](https://mlsanity.com/paper/2211.16566/) (2022)
- [Strategically Linked Decisions in Long-Term Planning and Reinforcement Learning](https://mlsanity.com/paper/2505.16833/) (2025)

## Abstract (arXiv)

We develop a Reinforcement Learning (RL) framework for improving an existing behavior policy via sparse, user-interpretable changes. Our goal is to make minimal changes while gaining as much benefit as possible. We define a minimal change as having a sparse, global contrastive explanation between the original and proposed policy. We improve the current policy with the constraint of keeping that global contrastive explanation short. We demonstrate our framework with a discrete MDP and a continuous 2D navigation domain.