ArticleProceedings of the National Academy of Sciences of the United States of America2024
Impossibility theorems for feature attribution.
Article in Proceedings of the National Academy of Sciences of the United States of America, 2024. The graph could read no effect estimate from its abstract, so it casts no vote on the map. Cited by 17 papers.
What it found
Each row is one number read from the abstract, on the scale the paper reported it, with its interval. Left of the dashed line favours the treatment, right favours the comparator. Under each row is the sentence it came from. New to these charts? A ten-minute tutorial.
The abstract states no effect estimate the extractor could read, or names no intervention and outcome on the map, so this paper lights no cell and moves no belief. It is still indexed, cited and linked below.
The trial behind it
Trials whose registry record cites this paper, or whose number appears in the abstract. A trial that started after this paper was published is citing it as background, not reporting it.
Neither the registry nor the abstract names a trial number. If this is a trial report, that itself is worth knowing.
Who cites it
17 citing papers in PubMed.
- Feature Importance Bias and Competing Risks: Methodological Concerns in Ensemble Models for Diabetic Kidney Disease.Diabetes, obesity & metabolism · 2026Article
- The most important features in generalized additive models might be groups of features.Scientific reports · 2026Article
- Explainable machine learning with routine biomarkers identifies culture-defined bacteremic urosepsis.Scientific reports · 2026Article
- Addressing biases and limitations in feature attribution for circRNA modification profiling.Briefings in bioinformatics · 2026Article
- Interpretable Cancer Survival Prediction by Fusing Semantic Labelling of Cell Types and Whole Slide Images.Interdisciplinary sciences, computational life sciences · 2026Article
- Mask of Truth: Model Sensitivity to Unexpected Regions of Medical Images.Journal of imaging informatics in medicine · 2026Article
- Feature representation for explainable CRISPR off-target prediction and base editing efficiency.Frontiers in bioinformatics · 2026Article
- Development and Validation of an Interpretable Machine Learning Model for Prediction of the Need for Surgical Evacuation in Patients with Incomplete Abortion.Risk management and healthcare policy · 2026Article
- Letter to the Editor: Navigating bias in machine learning-reevaluating feature importances through robust statistical analysis.European radiology · 2026Article
- Explainable AI needs formalization.NPJ artificial intelligence · 2026Review
- Bridging the human-AI knowledge gap through concept discovery and transfer in AlphaZero.Proceedings of the National Academy of Sciences of the United States of America · 2025Article
- Joint embedding-classifier learning for interpretable collaborative filtering.BMC bioinformatics · 2025Article
- Finding the needle in the haystack-An interpretable sequential pattern mining method for classification problems.Frontiers in big data · 2025Article
- Should Artificial Intelligence Play a Durable Role in Biomedical Research and Practice?International journal of molecular sciences · 2024Review
- Explainable AI for computational pathology identifies model limitations and tissue biomarkers.ArXiv · 2024Article
- Article
- MiMICRI: Towards Domain-centered Counterfactual Explanations of Cardiovascular Image Classification Models.Proceedings of the ... Conference on Fairness, Accountability, and Transparency · 2024Article
Corrections and comments
PubMed lists nothing against this paper. Absence here is not a guarantee, only a check that was made.
Authors and funding
4 authors.
Funding
No grant is acknowledged in the PubMed record.
Abstract
Despite a sea of interpretability methods that can produce plausible explanations, the field has also empirically seen many failure cases of such methods. In light of these results, it remains unclear for practitioners how to use these methods and choose between them in a principled way. In this paper, we show that for moderately rich model classes (easily satisfied by neural networks), any feature attribution method that is complete and linear-for example, Integrated Gradients and Shapley Additive Explanations (SHAP)-can provably fail to improve on random guessing for inferring model behavior. Our results apply to common end-tasks such as characterizing local model behavior, identifying spurious features, and algorithmic recourse. One takeaway from our work is the importance of concretely defining end-tasks: Once such an end-task is defined, a simple and direct approach of repeated model evaluations can outperform many other complex feature attribution methods.
Indexed as
Identifiers
What OpenQuestion holds
Registered trials
Read under generation 80e0d062 · epoch 390. Bibliography from PubMed, PubMed Central and OpenAlex; grants from NIH RePORTER; trial links from ClinicalTrials.gov; estimates, votes and beliefs from the OpenQuestion graph.