New SNRE framework offers regional explanations for ML models

Researchers have introduced a new framework called Causal Sufficient and Necessary Regional Explanations (SNRE) to characterize machine learning model predictions at a region level. This approach aims to identify input regions that are both sufficient and necessary for a model's output to fall within a specific range. SNRE utilizes a differentiable estimator derived from the Probability of Necessity and Sufficiency (PNS) and employs interpretable algebraic region families with learnable feature masks to balance expressiveness and interpretability. Experiments indicate that SNRE effectively learns region pairs demonstrating strong sufficiency-necessity performance, robust explanation behavior, and practical utility for model analysis. AI

IMPACT Introduces a novel method for regional model explainability, potentially improving trust and analysis of AI systems.

RANK_REASON This is a research paper published on arXiv detailing a new framework for machine learning explainability. [lever_c_demoted from research: ic=1 ai=1.0]

Read on arXiv cs.LG →

• alphaXiv
• arXiv
• CatalyzeX Code Finder for Papers
• Causal Sufficient and Necessary Regional Explanations
• CORE Recommender
• DagsHub
• Gotit.pub
• Hugging Face
• IArxiv Recommender
• Influence Flower
• Probability of Necessity and Sufficiency
• ScienceCast
• University of Michigan School for Environment and Sustainability

• paper
• other

AI-generated summary · Google Gemini · from 1 sources. How we write summaries →