ExplAIner Language Improves ML Model Explanation Queries

Marcelo Arenas, Pablo Barcel\'o, Diego Bustamante, Jose Caraball, Mar\'ia Alejandra Schild, Bernardo Subercaseaux· July 8, 2026 View original

▶ The 2-minute explainer

Summary

Researchers developed ExplAIner, a declarative query language for explaining classification models, addressing limitations of previous methods like FOIL. It can express various explanation notions and offers efficient evaluation for different Boolean models.

This research introduces ExplAIner, a novel declarative query language designed to enhance the explainability of machine learning classification models. It builds upon existing interpretability frameworks but overcomes significant limitations, particularly in expressing complex optimality-based explanation queries and improving computational efficiency. ExplAIner's layered structure and extended vocabulary allow it to articulate a wide range of explanation types, including abductive, contrastive, feature-based, and distance-based queries. The framework also demonstrates improved evaluation complexity, showing that queries can be processed with a fixed or polynomial number of calls to a SAT solver, making it practical for formal XAI applications.

Why it matters

Professionals can leverage this framework to more precisely define and efficiently compute explanations for their AI models, leading to greater transparency and trustworthiness in AI-driven decisions.

How to implement this in your domain

  1. 1Investigate ExplAIner's capabilities for specific model explanation needs.
  2. 2Integrate SAT solvers into existing MLOps pipelines for efficient explanation generation.
  3. 3Develop custom explanation queries using ExplAIner's declarative syntax.
  4. 4Evaluate the computational benefits of ExplAIner compared to current XAI methods.

Who benefits

HealthcareBFSIAutomotiveLegalCompliance

Key takeaways

  • ExplAIner offers a declarative language for comprehensive ML model explanations.
  • It addresses limitations of prior interpretability query languages.
  • The framework supports various explanation types, including abductive and contrastive.
  • Efficient evaluation is achieved through a fixed number of SAT solver calls.

Original post by Marcelo Arenas, Pablo Barcel\'o, Diego Bustamante, Jose Caraball, Mar\'ia Alejandra Schild, Bernardo Subercaseaux

"arXiv:2607.06407v1 Announce Type: new Abstract: The XAI community has studied a wide range of queries and scores for explaining predictions of ML models. From a data management perspective, this proliferation of explanation notions calls for declarative query languages in which s…"

View on X

Originally posted by Marcelo Arenas, Pablo Barcel\'o, Diego Bustamante, Jose Caraball, Mar\'ia Alejandra Schild, Bernardo Subercaseaux on X · view source

Want to go deeper?

Turn these trends into skills with Learnijoy's hands-on AI & tech courses.

Explore courses