The field of Explainable AI (XAI) and Interpretable Machine Learning (IML) is rapidly advancing, with a focus on developing techniques to provide insights into the decision-making processes of complex models. Recent research has explored various approaches, including feature attribution methods, concept-based explanations, and model-agnostic interpretability techniques. A key direction in the field is the development of methods that can provide human-understandable explanations, such as visualizations and natural language interpretations. Noteworthy papers in this area include FINCH, which introduces a visual analytics tool for explaining feature interactions in black box models, and TraNCE, which presents a transformative nonlinear concept explainer for CNNs. These advancements have the potential to increase trust and transparency in AI systems, enabling their deployment in high-stakes applications such as healthcare and cybersecurity.
Advances in Explainable AI and Interpretable Machine Learning
Sources
FINCH: Locally Visualizing Higher-Order Feature Interactions in Black Box Models
Towards Biomarker Discovery for Early Cerebral Palsy Detection: Evaluating Explanations Through Kinematic Perturbations
Interpretable Machine Learning for Oral Lesion Diagnosis through Prototypical Instances Identification
Exploring a Principled Framework for Deep Subspace Clustering
What's Producible May Not Be Reachable: Measuring the Steerability of Generative Models
Unraveling Pedestrian Fatality Patterns: A Comparative Study with Explainable AI
Z-REx: Human-Interpretable GNN Explanations for Real Estate Recommendations
Interpretable Feature Interaction via Statistical Self-supervised Learning on Tabular Data
Self-Explaining Neural Networks for Business Process Monitoring
Feature Learning beyond the Lazy-Rich Dichotomy: Insights from Representational Geometry
Exploring Energy Landscapes for Minimal Counterfactual Explanations: Applications in Cybersecurity and Beyond
Towards Human-Understandable Multi-Dimensional Concept Discovery
Interpretable Generative Models through Post-hoc Concept Bottlenecks
Extracting Interpretable Logic Rules from Graph Neural Networks
SMT-EX: An Explainable Surrogate Modeling Toolbox for Mixed-Variables Design Exploration
Guidelines For The Choice Of The Baseline in XAI Attribution Methods
Geometric Meta-Learning via Coupled Ricci Flow: Unifying Knowledge Representation and Quantum Entanglement
Random feature-based double Vovk-Azoury-Warmuth algorithm for online multi-kernel learning
TraNCE: Transformative Non-linear Concept Explainer for CNNs
Diffusion Counterfactuals for Image Regressors
MindfulLIME: A Stable Solution for Explanations of Machine Learning Models with Enhanced Localization Precision -- A Medical Image Case Study
EXPLICATE: Enhancing Phishing Detection through Explainable AI and LLM-Powered Interpretability
BioX-CPath: Biologically-driven Explainable Diagnostics for Multistain IHC Computational Pathology
Investigating the Duality of Interpretability and Explainability in Machine Learning
Consistent Multigroup Low-Rank Approximation
MASCOTS: Model-Agnostic Symbolic COunterfactual explanations for Time Series
VITAL: More Understandable Feature Visualization through Distribution Alignment and Relevant Information Flow
Learnable cut flow