Feature Attribution Stability Suite: How Stable Are Post-Hoc Attributions?
The George Washington University(乔治华盛顿大学)
Kamalasankari Subramaniakuppusamy, Jugal Gajjar
备注:
Accepted in the proceedings track of XAI4CV Workshop at CVPR 2026. It has 2 images, 5 tables, 6 equations, and 35 references in the main paper and 12 figures, 15 tables, and 3 references in the supplementary material
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.
Software for Dataset-wide XAI: From Local Explanations to Global Insights with Zennit, CoRelAy, and ViRelAy
Technische Universität Berlin(柏林工业大学);BIFOLD – Berlin Institute for the Foundations of Learning and Data(柏林学习与数据基础研究所);Fraunhofer Heinrich Hertz Institute(弗劳恩霍夫海因里希·赫兹研究所);Korea University(高丽大学);Max Planck Institut für Informatik(马克斯·普朗克信息学研究所)
Provably Better Explanations with Optimized Aggregation of Feature Attributions
LMU Munich(慕尼黑大学);Siemens AG(西门子公司);Technical University of Munich(慕尼黑工业大学);University of Oxford(牛津大学);Munich Center for Machine Learning (MCML)(慕尼黑机器学习中心);German Cancer Research Center (DKFZ)(德国癌症研究中心);Goethe University Frankfurt(法兰克福大学)
Revealing Hidden Context Bias in Segmentation and Object Detection through Concept-specific Explanations
Fraunhofer Heinrich-Hertz-Institute(弗劳恩霍夫海因里希·赫兹研究所);Technische Universität Berlin(柏林工业大学);Berlin Institute for the Foundations of Learning and Data(柏林学习与数据基础研究所)
Evaluating Post-hoc Interpretability with Intrinsic Interpretability
University of Coimbra(科英布拉大学);IPO-Porto Research Centre(波尔图肿瘤学研究所研究中心);University of Applied Sciences Western Switzerland (HES-SO Valais)(瑞士西部应用科学大学(瓦莱州高等专业学院))
On Thin Ice: Towards Explainable Conservation Monitoring via Attribution and Perturbations
Duke University(杜克大学);University of Agder(阿格德大学);University of Cambridge(剑桥大学);U.S. National Park Service Department of Interior(美国国家公园管理局);Alaska Spatial Science Fairbanks, AK(阿拉斯加空间科学研究所)
University of Tübingen(蒂宾根大学);Sogang University(西江大学);Max Planck Institute for Intelligent Systems(马克斯·普朗克智能系统研究所);Max Planck Institute for Informatics(马克斯·普朗克信息学研究所);NAVER AI Lab(NAVER人工智能实验室)
AURA: Active-Response Attribution under Treatment Ambiguity in Bacterial Cytological Profiling
Nanyang Technological University(南洋理工大学);Institute of High Performance Computing, A*STAR(新加坡科技研究局高性能计算研究所);University of California, Santa Barbara(加州大学圣塔芭芭拉分校)