Back to feed
AR
arXiv CS.AI
7/17/2026
The original title is "IMEX Interaction-Based Model Explanation" which is quite short (5 words). Let me check the rules:

The original title is "IMEX Interaction-Based Model Explanation" which is quite short (5 words). Let me check the rules:

Original: IMEX Interaction-Based Model Explanation

Short summary

IMEX is an explainable AI method that identifies which variables and variable interactions contribute most to a model's predictions, supporting higher-order interaction analysis beyond pairwise effects. It uses two metrics: Static Correlation Power for individual feature contributions and Interaction Correlation Power for non-additive effects. Experimental validation against INVASE on three synthetic datasets shows IMEX recovers relevant feature structures even under non-linear, conditional, and multicollinear relationships.

  • •IMEX identifies feature importance and higher-order interactions in black-box models
  • •Two metrics (PCS and PCI) quantify individual and interaction contributions respectively
  • •Validated against INVASE, recovering feature structures under non-linear and multicollinear conditions

Generated with AI, which can make mistakes.

#research-breakthrough
Read full article at arXiv CS.AI

Is this a good recommendation for you?

Comments

Failed to load comments. Please try again.

Explore more

CIExplainer++: Generating Causal and Interpretable Explanations for Graph Neural Networks
arXiv cs.LG
Towards the Explainability of Temporal Graph Networks via Memory Backtracking and Topological Attribution
arXiv cs.LG
Position: Explainability Research Must Prioritize Foundations over Ad-hoc Methods
arXiv cs.LG
From Graphs to Gradients: Physics-Inspired Structural Attribution for Cyber-Physical IoT Systems and Beyond
arXiv CS.AI