Research map / Label-efficient and robust learning / Recognition and long-tail
Regularizing Black-box Models for Improved Interpretability
Gregory Plumb, Maruan Al-Shedivat, Ángel Alexander Cabrera, Adam Perer, Eric P. Xing, Ameet Talwalkar
Read the paperCheck your idea against 137,217 papers
AI-extracted. The measured effects below were read by AI from the paper's full text (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 |
|---|---|---|---|
| add ExpO fidelity regularizer to training | MLP regression models · UCI regression datasets (LIME/MAPLE) | interpretability metrics (PF, NF, stability): e.g. winequality LIME-NF 0.0330 -> 0.0025 | Improved |
| add ExpO fidelity regularizer | MLP regression models · communities, day, MSD | test MSE: communities 0.49 -> 0.46; MSD 0.583 -> 0.557 | Improved |
| use 1D-fidelity regularizer instead of full fidelity regularizer (ablation) | ExpO · UCI regression datasets | interpretability metrics: autompgs LIME-NF 0.011 -> 0.029 (None 0.041) | Mixed |
| add ExpO fidelity regularizer to compare with SENN (tuned for accuracy) | breast cancer classifier · breast cancer | accuracy: 0.92 (SENN) -> 0.99 | Improved |
| add l1 or l2 regularization instead of ExpO (ablation) | MLP regression models · UCI regression datasets | interpretability metrics: housing LIME-NF 0.14 -> 0.13 (L1), 0.12 (L2) | No clear effect |
| ExpO-regularized model vs normally trained model in human user study | explanations for housing model · user study | steps to reach target: 11.45 -> 8.00 | Improved |
| add ExpO fidelity regularizer on support2 mortality classification | MLP classifier · SUPPORT2 | accuracy | No clear effect |
Closest papers by meaning
- Towards Robust Interpretability with Self-Explaining Neural Networks (NeurIPS 2018 · similarity 0.88)
- Enhancing Performance of Explainable AI Models with Constrained Concept Refinement (ICML 2025 · similarity 0.87)
- Automated Assessment of Fidelity and Interpretability: An Evaluation Framework for Large Language Models’ Explanations (Student Abstract) (AAAI 2024 · similarity 0.86)
- Explainable Neural Networks with Guarantee: A Sparse Estimation Approach (AAAI 2025 · similarity 0.85)
- Predictive Concept Decoders: Training Scalable End-to-End Interpretability Assistants (NeurIPS 2026 · similarity 0.85)
- Local vs. Global Interpretability: A Computational Complexity Perspective (ICML 2024 · similarity 0.85)
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.
