Research map / Learning theory and algorithms / Prediction and decision losses
Model Agnostic Multilevel Explanations
Read the paperCheck your idea against 108,967 papers
AI-extracted summary. Written by AI from the paper's title only (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Measured effects reported in the paper
| What was changed | Where | Measured | Effect |
|---|---|---|---|
| Use MAME instead of LIME for multilevel feature explanations | Random Forest explanations · Auto MPG, Retention, HELOC, and ATIS | Kendall's rank correlation: 0.4800 -> 0.5267; 0.5158 -> 0.5474; 0.6623 -> 0.6883; 0.6976 -> 0.7547 | Helped |
| Use MAME instead of Two Step for multilevel feature explanations | Random Forest explanations · Auto MPG, Retention, HELOC, and ATIS | Kendall's rank correlation: 0.4667 -> 0.5267; 0.5368 -> 0.5474; 0.6537 -> 0.6883; 0.7089 -> 0.7547 | Helped |
| Use MAME instead of Two Step for multilevel feature explanations | Random Forest explanations · Waveform | Kendall's rank correlation: 0.5333 -> 0.2952 | Hurt |
| Use MAME rather than Two Step or SP-LIME to provide a small number of high-level clusters | MLP black-box explanations · Classification datasets with few clusters | Generalized fidelity | Helped |
| Use MAME rather than Two Step or SP-LIME to provide high-level explanations | HELOC loan prediction user study · HELOC user study | Participant prediction MSE: 0.036 | Helped |
| Use MAME rather than Two Step or SP-LIME to provide high-level explanations | HELOC loan prediction user study · HELOC user study | Correct most-similar cluster choices: 7 out of 10; less than 3 correct clusters | Helped |
| Use the AR-based approximate optimization rather than repeatedly solving each beta value to convergence | MAME · Auto MPG | Runtime: around 10 times faster | Helped |
Closest papers by meaning
- Trust Regions for Explanations via Black-Box Probabilistic Certification (ICML 2024 · similarity 0.84)
- Stratified GNN Explanations through Sufficient Expansion (AAAI 2024 · similarity 0.84)
- Local vs. Global Interpretability: A Computational Complexity Perspective (ICML 2024 · similarity 0.84)
- CohEx: A Generalized Framework for Cohort Explanation (AAAI 2025 · similarity 0.84)
- Practical do-Shapley Explanations with Estimand-Agnostic Causal Inference (NeurIPS 2025 · similarity 0.84)
- SpEx: A Spectral Approach to Explainable Clustering (NeurIPS 2025 · similarity 0.83)
Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.