Explaining Explanations: An Overview of Interpretability of Machine\n Learning

2018/05/31 by Leilani H. Gilpin, Gilpin, Leilani H., David Bau +9 · 37 citations
Computer Science · #Explainable Artificial Intelligence (XAI) #Adversarial Robustness in Machine Learning #Anomaly Detection Techniques and Applications

paper · pdf · doi:10.48550/arxiv.1806.00069

Abstract

There has recently been a surge of work in explanatory artificial\nintelligence (XAI). This research area tackles the important problem that\ncomplex machines and algorithms often cannot provide insights into their\nbehavior and thought processes. XAI allows users and parts of the internal\nsystem to be more transparent, providing explanations of their decisions in\nsome level of detail. These explanations are important to ensure algorithmic\nfairness, identify potential bias/problems in the training data, and to ensure\nthat the algorithms perform as expected. However, explanations produced by\nthese systems is neither standardized nor systematically assessed. In an effort\nto create best practices and identify open challenges, we provide our\ndefinition of explainability and show how it can be used to classify existing\nliterature. We discuss why current approaches to explanatory methods especially\nfor deep neural networks are insufficient. Finally, based on our survey, we\nconclude with suggested future research directions for explanatory artificial\nintelligence.\n

Cited by

• Block-Recurrent Dynamics in Vision Transformers
• Self-organizing maps for water quality assessment in reservoirs and lakes: A systematic literature review
• Explainable AI in Big Data Fraud Detection
• A Geometric Unification of Concept Learning with Concept Cones
• Modular Jets for Supervised Pipelines: Diagnosing Mirage vs Identifiability
• Sample-Efficient Counterfactual Tuning for Compressor Pressure Control
• Guaranteed Optimal Compositional Explanations for Neurons
• Open Vocabulary Compositional Explanations for Neuron Alignment
• White-Box Modeling of V2X Link Performance Using Stabilized Symbolic Regression
• Rethinking Saliency Maps: A Cognitive Human Aligned Taxonomy and Evaluation Framework for Explanations
• Towards Strong Certified Defense with Universal Asymmetric Randomization
• Automatic Piecewise Linear Regression for Predicting Student Learning Satisfaction
• Utilizing Large Language Models for Machine Learning Explainability
• Cluster Paths: Navigating Interpretability in Neural Networks
• TreeNet: Layered Decision Ensembles
• Memory-Augmented Log Analysis with Phi-4-mini: Enhancing Threat Detection in Structured Security Logs
• ACE: Adapting sampling for Counterfactual Explanations
• EXP-CAM: Explanation Generation and Circuit Discovery Using Classifier Activation Matching
• AutoML in Cybersecurity: An Empirical Study
• Activation Matching for Explanation Generation
• Advances and Challenges in Machine Learning‐based Image Analysis for Monitoring and Predicting Organic Crystal Formation
• Towards Transparent AI: A Survey on Explainable Language Models
• A Comparative Study of Rule-Based and Data-Driven Approaches in Industrial Monitoring
• A Framework for Generating Artificial Datasets to Validate Absolute and Relative Position Concepts
• An Interpretable Deep Learning Model for General Insurance Pricing
• An Approach to Grounding AI Model Evaluations in Human-derived Criteria
• Towards LLM-generated explanations for Component-based Knowledge Graph Question Answering Systems
• Checkmate: interpretable and explainable RSVQA is the endgame
• From Black Box to Transparency: Enhancing Automated Interpreting Assessment with Explainable AI in College Classrooms
• Exploring Content and Social Connections of Fake News with Explainable Text and Graph Learning
• Towards Transparent Ethical AI: A Roadmap for Trustworthy Robotic Systems
• Decoding the Multimodal Maze: A Systematic Review on the Adoption of Explainability in Multimodal Attention-based Models
• Error Detection and Correction for Interpretable Mathematics in Large Language Models
• Probing the Gaps in ChatGPT Live Video Chat for Real-World Assistance for People who are Blind or Visually Impaired
• Toward using explainable data-driven surrogate models for treating performance-based seismic design as an inverse engineering problem
• Explainability Through Systematicity: The Hard Systematicity Challenge for Artificial Intelligence
• Finding Uncommon Ground: A Human-Centered Model for Extrospective Explanations

Related

• Axe the X in XAI: A Plea for Understandable AI
• Explaining Machine Learning Decisions
• "Explanation" is Not a Technical Term: The Problem of Ambiguity in XAI
• How do Humans Understand Explanations from Machine Learning Systems? An Evaluation of the Human-Interpretability of Explanation
• Explaining Explainability: Towards Deeper Actionable Insights into Deep Learning through Second-order Explainability