Interpretable Representations in Explainable AI: From Theory to Practice
Preprint
Abstract
Interpretable representations are the backbone of many explainers that target black-box predictive systems based on artificial intelligence and machine learning algorithms. They translate the low-level data representation necessary for good predictive performance into high-level human-intelligible concepts used to convey the explanatory insights. Notably, the explanation type and its cognitive complexity are directly controlled by the interpretable representation, tweaking which allows to target a particular audience and use case. However, many explainers built upon interpretable representations overlook their merit and fall back on default solutions that often carry implicit assumptions, thereby degrading the explanatory power and reliability of such techniques. To address this problem, we study properties of interpretable representations that encode presence and absence of human-comprehensible concepts. We demonstrate how they are operationalised for tabular, image and text data; discuss their assumptions, strengths and weaknesses; identify their core building blocks; and scrutinise their configuration and parameterisation. In particular, this in-depth analysis allows us to pinpoint their explanatory properties, desiderata and scope for (malicious) manipulation in the context of tabular data where a linear model is used to quantify the influence of interpretable concepts on a black-box prediction. Our findings lead to a range of recommendations for designing trustworthy interpretable representations; specifically, the benefits of class-aware (supervised) discretisation of tabular data, e.g., with decision trees, and sensitivity of image interpretable representations to segmentation granularity and occlusion colour.Keywords
All Related Versions
This publication has 8 references indexed in Scilit:
- LIME: Consistent and Faithful Surrogate Explanations of Multiple ClassesElectronics, 2025
- What and How of Machine Learning Transparency: Building Bespoke Explainability Tools with Interoperable Algorithmic ComponentsJournal of Open Source Education, 2022
- FAT Forensics: A Python toolbox for algorithmic fairness, accountability and transparencySoftware Impacts, 2022
- FAT Forensics: A Python Toolbox for Implementing and Deploying Fairness, Accountability and Transparency Algorithms in Predictive SystemsJournal of Open Source Software, 2020
- Stop explaining black box machine learning models for high stakes decisions and use interpretable models insteadNature Machine Intelligence, 2019
- Glass-Box: Explaining AI Decisions With Counterfactual Statements Through Conversation With a Voice-enabled Virtual Assistant2018
- scikit-image: image processing in PythonPeerJ, 2014
- SLIC Superpixels Compared to State-of-the-Art Superpixel MethodsIEEE Transactions on Pattern Analysis and Machine Intelligence, 2012