What Does Explainable AI Really Mean? A New Conceptualization of Perspectives

Abstract

We characterize three notions of explainable AI that cut across research fields: opaque systems that offer no insight into its algo- rithmic mechanisms; interpretable systems where users can mathemat- ically analyze its algorithmic mechanisms; and comprehensible systems that emit symbols enabling user-driven explanations of how a conclusion is reached. The paper is motivated by a corpus analysis of NIPS, ACL, COGSCI, and ICCV/ECCV paper titles showing differences in how work on explainable AI is positioned in various fields. We close by introducing a fourth notion: truly explainable systems, where automated reasoning is central to output crafted explanations without requiring human post processing as final step of the generative process.

References in corpus (2)

• Image Captioning with Semantic Attention
• Can you tell where in India I am from? Comparing humans and computers on fine-grained race face classification

Cited by in corpus (43)

• A Survey on the Explainability of Supervised Machine Learning
• The role of explainability in creating trustworthy artificial intelligence for health care: a comprehensive survey of the terminology, design choices, and evaluation strategies
• What do we need to build explainable AI systems for the medical domain?
• What Do We Want From Explainable Artificial Intelligence (XAI)? -- A Stakeholder Perspective on XAI and a Conceptual Model Guiding Interdisciplinary XAI Research
• Opportunities and Challenges in Explainable Artificial Intelligence (XAI): A Survey
• Explanation in Human-AI Systems: A Literature Meta-Review, Synopsis of Key Ideas and Publications, and Bibliography for Explainable AI
• Explainable Artificial Intelligence: A Survey of Needs, Techniques, Applications, and Future Direction
• Trustworthy AI and Robotics and the Implications for the AEC Industry: A Systematic Literature Review and Future Potentials
• Explainable Artificial Intelligence (XAI) on TimeSeries Data: A Survey
• Robust Explainability: A Tutorial on Gradient-Based Attribution Methods for Deep Neural Networks
• Explainable Artificial Intelligence and Machine Learning: A reality rooted perspective
• To trust or not to trust an explanation: using LEAF to evaluate local linear XAI methods
• A Manifesto for Future Generation Cloud Computing: Research Directions for the Next Decade
• `It is currently hodgepodge'': Examining AI/ML Practitioners' Challenges during Co-production of Responsible AI Values
• Catala: A Programming Language for the Law
• Discovering interpretable elastoplasticity models via the neural polynomial method enabled symbolic regressions
• Explainable Artificial Intelligence (XAI): An Engineering Perspective
• Defining Explainable AI for Requirements Analysis
• Responsible AI by Design in Practice
• Towards Explainable Neural-Symbolic Visual Reasoning
• Explainable Machine Learning with Prior Knowledge: An Overview
• Stacked Penalized Logistic Regression for Selecting Views in Multi-View Learning
• User-centered Feedback Design in Person-following Robots for Older Adults
• Exploiting auto-encoders and segmentation methods for middle-level explanations of image classification systems
• Interpreting Embedding Models of Knowledge Bases: A Pedagogical Approach
• Generating User-friendly Explanations for Loan Denials using GANs
• The What, the Why, and the How of Artificial Explanations in Automated Decision-Making
• Relating Input Concepts to Convolutional Neural Network Decisions
• Collaboratively boosting data-driven deep learning and knowledge-guided ontological reasoning for semantic segmentation of remote sensing imagery
• Sequential Explanations with Mental Model-Based Policies
• Explainable AI and Adoption of Financial Algorithmic Advisors: an Experimental Study
• Quantifying the Task-Specific Information in Text-Based Classifications
• A Gaze into the Internal Logic of Graph Neural Networks, with Logic
• Ten AI Stepping Stones for Cybersecurity
• Economics of Human-AI Ecosystem: Value Bias and Lost Utility in Multi-Dimensional Gaps
• Explainable AI for Classification using Probabilistic Logic Inference
• Teaching Solid Mechanics to Artificial Intelligence: a fast solver for heterogeneous solids
• Generalized Constraints as A New Mathematical Problem in Artificial Intelligence: A Review and Perspective
• XAlgo: a Design Probe of Explaining Algorithms' Internal States via Question-Answering
• Scientific Calculator for Designing Trojan Detectors in Neural Networks
• Semantic bottleneck for computer vision tasks
• Non-Asimov Explanations Regulating AI through Transparency
• Memory networks for consumer protection:unfairness exposed