Feature Attribution Stability Suite: How Stable Are Post-Hoc Attributions?

• URL: http://arxiv.org/abs/2604.02532v1
• Date: Thu, 02 Apr 2026 21:32:54 GMT
• Title: Feature Attribution Stability Suite: How Stable Are Post-Hoc Attributions?
• Abstract summary: Post-hoc feature attribution methods are widely deployed in safety-critical vision systems.<n>We introduce the Feature Stability Suite (FASS), a benchmark that enforces prediction-invariance filtering.<n>FASS decomposes stability into three complementary metrics: structural similarity, rank correlation, and top-k Jaccard overlap.
• Score: 0.0
• License: http://creativecommons.org/licenses/by-nc-sa/4.0/
• Abstract: Post-hoc feature attribution methods are widely deployed in safety-critical vision systems, yet their stability under realistic input perturbations remains poorly characterized. Existing metrics evaluate explanations primarily under additive noise, collapse stability to a single scalar, and fail to condition on prediction preservation, conflating explanation fragility with model sensitivity. We introduce the Feature Attribution Stability Suite (FASS), a benchmark that enforces prediction-invariance filtering, decomposes stability into three complementary metrics: structural similarity, rank correlation, and top-k Jaccard overlap-and evaluates across geometric, photometric, and compression perturbations. Evaluating four attribution methods (Integrated Gradients, GradientSHAP, Grad-CAM, LIME) across four architectures and three datasets-ImageNet-1K, MS COCO, and CIFAR-10, FASS shows that stability estimates depend critically on perturbation family and prediction-invariance filtering. Geometric perturbations expose substantially greater attribution instability than photometric changes, and without conditioning on prediction preservation, up to 99% of evaluated pairs involve changed predictions. Under this controlled evaluation, we observe consistent method-level trends, with Grad-CAM achieving the highest stability across datasets.

Related papers

• The Road Taken: The Role of Optimizers at the Edge of Stability [53.81268610971847]
  edge of stability refers to a phenomenon in deep learning where the Hessian eigenvalues of the loss remain stable above a threshold that the classical descent predicts to be unstable.<n>Previous works formulate the edge of stability with respect to the maximum Hessian eigenvalue and the learning rate.<n>We observe that many first-order methods, including gradient descent, significantly violate the stability bound predicted by these theories by a factor as large as $times 21.1$.
  arXiv  Detail & Related papers  (2026-08-19T01:03:34Z)
• Conformal Changepoint Localization and Root Cause Analysis with Corrupted Observations [55.76952683833966]
  This paper proposes weighted CONCH (W-CONCH) and weighted CROC (W-CROC) to reduce confidence set size when data may be corrupted.<n>Experiments on image-based and real-world changepoint and root-cause benchmarks show that uncertainty-based weighting substantially reduces confidence set size.
  arXiv  Detail & Related papers  (2026-07-29T05:16:59Z)
• Reliability-Aware Prototype Calibration for Frozen Pose-Flow Video Anomaly Detection [7.276043252948729]
  A single likelihood score may hide normal behavior and be sensitive to pose-observation noise.<n>We study a frozen-detector setting in which the pose-flow backbone, cached skeleton tracks, and evaluation pipeline are fixed.<n> Reliability-Aware Prototype is a post-hoc score calibration method for this setting.
  arXiv  Detail & Related papers  (2026-06-18T14:46:38Z)
• Link Prediction or Perdition: the Seeds of Instability in Knowledge Graph Embeddings [2.1174215880331766]
  Embedding models (KGEMs) constitute the main link prediction approach to complete knowledge graphs.<n>Standard evaluation protocols emphasize rank-based metrics such as MRR or Hits@$K$, but usually overlook the influence of random seeds on result stability.<n>We find that high-performance models actually produce divergent predictions at the triple level and highly variable spaces.
  arXiv  Detail & Related papers  (2026-06-02T09:13:50Z)
• Fortress: A Case Study in Stabilizing Search Recommendations via Temporal Data Augmentation and Feature Pruning [2.6494136490021387]
  Fortress identifies and prunes features that contribute to inconsistent prediction scores over time.<n>We validate Fortress on a query-to-app relevance model in a large-scale app marketplace.
  arXiv  Detail & Related papers  (2026-05-14T18:13:05Z)
• Multimodal Learning on Low-Quality Data with Conformal Predictive Self-Calibration [72.0672328514289]
  Multimodal learning often grapples with the challenge of low-quality data.<n>We propose a unified framework, termed Conformal Predictive Self-Calibration.<n>Our framework consistently outperforms existing state-of-the-art methods.
  arXiv  Detail & Related papers  (2026-05-05T14:48:52Z)
• SAIF: A Stability-Aware Inference Framework for Medical Image Segmentation with Segment Anything Model [11.410993498576936]
  Segment Anything Model (SAM) enable scalable medical image segmentation but suffer from inference-time instability when deployed as a frozen backbone.<n>We propose the Stability-Aware Inference Framework (SAIF), a training-free and plug-and-play inference framework that improves robustness by explicitly modeling prompt and threshold uncertainty.
  arXiv  Detail & Related papers  (2026-03-13T19:15:34Z)
• Measuring the Fragility of Trust: Devising Credibility Index via Explanation Stability (CIES) for Business Decision Support Systems [3.8615905456206256]
  This paper introduces the Credibility Index via Explanation Stability (CIES), a metric that measures how robust a model's explanations are when subject to realistic business noise.<n>CIES captures whether the reasons behind a prediction remain consistent, not just the prediction itself.<n>Results demonstrate that model complexity impacts explanation credibility, class imbalance treatment via SMOTE affects not only predictive performance but also explanation stability.
  arXiv  Detail & Related papers  (2026-03-05T10:11:55Z)
• On the Structural Non-Preservation of Epistemic Behaviour under Policy Transformation [51.56484100374058]
  We formalise such information-conditioned interaction patterns as behavioural dependency.<n>This induces a probe-relative notion of $$-behavioural equivalence and a within-policy behavioural distance.<n>Results identify structural conditions under which probe-conditioned behavioural separation is not preserved under common policy transformations.
  arXiv  Detail & Related papers  (2026-02-24T22:55:21Z)
• Not All Preferences Are Created Equal: Stability-Aware and Gradient-Efficient Alignment for Reasoning Models [52.48582333951919]
  We propose a dynamic framework designed to enhance alignment reliability by maximizing the Signal-to-Noise Ratio of policy updates.<n>SAGE (Stability-Aware Gradient Efficiency) integrates a coarse-grained curriculum mechanism that refreshes candidate pools based on model competence.<n> Experiments on multiple mathematical reasoning benchmarks demonstrate that SAGE significantly accelerates convergence and outperforms static baselines.
  arXiv  Detail & Related papers  (2026-02-01T12:56:10Z)
• Geometric Stability: The Missing Axis of Representations [0.0]
  We introduce $geometric$ $stability, a distinct dimension that quantifies how reliably representational geometry holds under perturbation.<n>Across 2,463 configurations in seven domains, we show that stability and similarity are empirically uncorrelated.<n>For safety monitoring, stability acts as a functional geometric canary, detecting structural drift nearly 2$times$ more sensitively than CKA.<n>For model selection, stability dissociates from transferability, revealing a geometric tax that transfer optimization incurs.
  arXiv  Detail & Related papers  (2026-01-14T05:15:22Z)
• MaP: A Unified Framework for Reliable Evaluation of Pre-training Dynamics [72.00014675808228]
  Instability in Large Language Models evaluation process obscures true learning dynamics.<n>We introduce textbfMaP, a framework that integrates underlineMerging underlineand the underlinePass@k metric.<n>Experiments show that MaP yields significantly smoother performance curves, reduces inter-run variance, and ensures more consistent rankings.
  arXiv  Detail & Related papers  (2025-10-10T11:40:27Z)
• Stable Update of Regression Trees [0.0]
  We focus on the stability of an inherently explainable machine learning method, namely regression trees. We propose a regularization method, where data points are weighted based on the uncertainty in the initial model. Results show that the proposed update method improves stability while achieving similar or better predictive performance.
  arXiv  Detail & Related papers  (2024-02-21T09:41:56Z)
• Evaluating Probabilistic Classifiers: The Triptych [62.997667081978825]
  We propose and study a triptych of diagnostic graphics that focus on distinct and complementary aspects of forecast performance. The reliability diagram addresses calibration, the receiver operating characteristic (ROC) curve diagnoses discrimination ability, and the Murphy diagram visualizes overall predictive performance and value.
  arXiv  Detail & Related papers  (2023-01-25T19:35:23Z)
• Evaluating probabilistic classifiers: Reliability diagrams and score decompositions revisited [68.8204255655161]
  We introduce the CORP approach, which generates provably statistically Consistent, Optimally binned, and Reproducible reliability diagrams in an automated way. Corpor is based on non-parametric isotonic regression and implemented via the Pool-adjacent-violators (PAV) algorithm.
  arXiv  Detail & Related papers  (2020-08-07T08:22:26Z)
• Fine-Grained Analysis of Stability and Generalization for Stochastic Gradient Descent [55.85456985750134]
  We introduce a new stability measure called on-average model stability, for which we develop novel bounds controlled by the risks of SGD iterates. This yields generalization bounds depending on the behavior of the best model, and leads to the first-ever-known fast bounds in the low-noise setting. To our best knowledge, this gives the firstever-known stability and generalization for SGD with even non-differentiable loss functions.
  arXiv  Detail & Related papers  (2020-06-15T06:30:19Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.