{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,2,11]],"date-time":"2026-02-11T13:47:05Z","timestamp":1770817625292,"version":"3.50.1"},"reference-count":34,"publisher":"Frontiers Media SA","license":[{"start":{"date-parts":[[2026,2,11]],"date-time":"2026-02-11T00:00:00Z","timestamp":1770768000000},"content-version":"vor","delay-in-days":0,"URL":"https:\/\/creativecommons.org\/licenses\/by\/4.0\/"}],"content-domain":{"domain":["frontiersin.org"],"crossmark-restriction":true},"short-container-title":["Front. Artif. Intell."],"abstract":"<jats:p>This work presents a conceptual framework for causal concept-based post-hoc explainable artificial intelligence (XAI), based on the requirements that explanations for non-interpretable models must be both understandable and faithful to the model being explained. Local and global explanations are generated by calculating the probability of sufficiency of concept interventions. Example explanations are presented, generated with a proof-of-concept model made to explain classifiers trained on the CelebA dataset. Understandability is demonstrated through a clear concept-based vocabulary, subject to an implicit causal interpretation. Fidelity is addressed by highlighting important framework assumptions, stressing that the context of explanation interpretation must align with the context of explanation generation.<\/jats:p>","DOI":"10.3389\/frai.2025.1759000","type":"journal-article","created":{"date-parts":[[2026,2,11]],"date-time":"2026-02-11T11:22:26Z","timestamp":1770808946000},"update-policy":"https:\/\/doi.org\/10.3389\/crossmark-policy","source":"Crossref","is-referenced-by-count":0,"title":["A framework for causal concept-based model explanations"],"prefix":"10.3389","volume":"8","author":[{"given":"Anna Rodum","family":"Bj\u00f8ru","sequence":"first","affiliation":[{"name":"Department of Computer Science, Norwegian University of Science and Technology","place":["Trondheim, Norway"]}]},{"given":"Jacob","family":"Lysn\u00e6s-Larsen","sequence":"additional","affiliation":[{"name":"Department of Computer Science, Norwegian University of Science and Technology","place":["Trondheim, Norway"]}]},{"given":"Oskar","family":"J\u00f8rgensen","sequence":"additional","affiliation":[{"name":"Department of Computer Science, Norwegian University of Science and Technology","place":["Trondheim, Norway"]}]},{"given":"Inga","family":"Str\u00fcmke","sequence":"additional","affiliation":[{"name":"Department of Computer Science, Norwegian University of Science and Technology","place":["Trondheim, Norway"]}]},{"given":"Helge","family":"Langseth","sequence":"additional","affiliation":[{"name":"Department of Computer Science, Norwegian University of Science and Technology","place":["Trondheim, Norway"]}]}],"member":"1965","published-online":{"date-parts":[[2026,2,11]]},"reference":[{"key":"B1","first-page":"9525","article-title":"\u201cSanity checks for saliency maps,\u201d","volume-title":"Proceedings of the International Conference on Neural Information Processing Systems","author":"Adebayo","year":"2018"},{"key":"B2","doi-asserted-by":"publisher","first-page":"e1424","DOI":"10.1002\/widm.1424","article-title":"Explainable artificial intelligence: an analytical review","volume":"11","author":"Angelov","year":"2021","journal-title":"WIREs Data Mining Knowl. Discov"},{"key":"B3","first-page":"1","article-title":"pgmpy: A Python toolkit for Bayesian networks","volume":"25","author":"Ankan","year":"2024","journal-title":"J. Mach. Learn. Res"},{"key":"B4","first-page":"357","article-title":"\u201cIdentifiability of path-specific effects,\u201d","volume-title":"Proceedings of the International Joint Conference on Artificial Intelligence","author":"Avin","year":"2005"},{"key":"B5","doi-asserted-by":"publisher","first-page":"82","DOI":"10.1016\/j.inffus.2019.12.012","article-title":"Explainable artificial intelligence (XAI): concepts, taxonomies, opportunities and challenges toward responsible AI","volume":"58","author":"Barredo Arrieta","year":"2020","journal-title":"Inform. Fusion"},{"key":"B6","first-page":"90","article-title":"\u201cCausal explanations and XAI,\u201d","volume-title":"Proceedings of the First Conference on Causal Learning and Reasoning","author":"Beckers","year":"2022"},{"key":"B7","doi-asserted-by":"publisher","first-page":"109520","DOI":"10.1016\/j.ijar.2025.109520","article-title":"Divide and conquer for causal computation","volume":"186","author":"Bj\u00f8ru","year":"","journal-title":"Int. J. Approxim. Reason"},{"key":"B8","doi-asserted-by":"publisher","DOI":"10.36227\/techrxiv.176945882.23563107\/v1","article-title":"Causal post-hoc XAI: categorization and systematic literature review","author":"Bj\u00f8ru","year":"","journal-title":"TechRxiv"},{"key":"B9","author":"Caba\u00f1as","year":"2025","journal-title":"BCAUSE: A Python Package for Causal and Counterfactual Reasoning Using Probabilistic Graphical models"},{"key":"B10","first-page":"8789","article-title":"\u201cStarGAN: Unified generative adversarial networks for multi-domain image-to-image translation,\u201d","volume-title":"Proceedings of the IEEE\/CVF Conference on Computer Vision and Pattern Recognition","author":"Choi","year":"2018"},{"key":"B11","doi-asserted-by":"publisher","DOI":"10.48550\/arXiv.2309.08600","article-title":"Sparse autoencoders find highly interpretable features in language models","author":"Cunningham","year":"2023","journal-title":"arXiv"},{"key":"B12","first-page":"3879","article-title":"\u201cEvaluating and mitigating bias in image classifiers: A causal perspective using counterfactuals,\u201d","volume-title":"Proceedings of the IEEE\/CVF Winter Conference on Applications of Computer Vision","author":"Dash","year":"2022"},{"key":"B13","first-page":"577","article-title":"\u201cExplaining black-box algorithms using probabilistic contrastive counterfactuals,\u201d","volume-title":"Proceedings of the ACM SIGMOD International Conference on Management of Data","author":"Galhotra","year":"2021"},{"key":"B14","doi-asserted-by":"publisher","first-page":"9","DOI":"10.1016\/S0004-3702(97)00047-7","article-title":"Axioms of causal relevance","volume":"97","author":"Galles","year":"1997","journal-title":"Artif. Intell"},{"key":"B15","doi-asserted-by":"crossref","DOI":"10.24251\/HICSS.2021.156","article-title":"\u201cReviewing the need for explainable artificial intelligence (xAI),\u201d","author":"Gerlings","year":"2021","journal-title":"Proceedings of the Hawaii International Conference on System Sciences"},{"key":"B16","unstructured":"Gildenblat\n              J.\n            \n          \n          PyTorch Library for CAM Methods\n          \n          2021"},{"key":"B17","doi-asserted-by":"publisher","first-page":"50","DOI":"10.1609\/aimag.v38i3.2741","article-title":"European union regulations on algorithmic decision-making and a \u201cright to explanation\u201d","volume":"38","author":"Goodman","year":"2017","journal-title":"AI Magazine"},{"key":"B18","doi-asserted-by":"publisher","first-page":"44","DOI":"10.1609\/aimag.v40i2.2850","article-title":"DARPA's explainable artificial intelligence program","volume":"40","author":"Gunning","year":"2019","journal-title":"AI Magazine"},{"key":"B19","first-page":"770","article-title":"\u201cDeep residual learning for image recognition,\u201d","volume-title":"Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition","author":"He","year":"2016"},{"key":"B20","first-page":"4198","article-title":"\u201cTowards faithfully interpretable NLP systems: How should we define and evaluate faithfulness?,\u201d","volume-title":"Proceedings of the Annual Meeting of the Association for Computational Linguistics","author":"Jacovi","year":"2020"},{"key":"B21","first-page":"2668","article-title":"\u201cInterpretability beyond feature attribution: Quantitative testing with concept activation vectors (TCAV),\u201d","volume-title":"International Conference on Machine Learning","author":"Kim","year":"2018"},{"key":"B22","doi-asserted-by":"publisher","first-page":"2516","DOI":"10.3233\/FAIA240780","article-title":"Causal diffusion autoencoders: Toward counterfactual generation via diffusion probabilistic models","volume":"392","author":"Komanduri","year":"2024","journal-title":"Front. Artif. Intellig. Appl"},{"key":"B23","article-title":"\u201cDeep learning face attributes in the wild,\u201d","author":"Liu","year":"2015","journal-title":"Proceedings of International Conference on Computer Vision (ICCV"},{"key":"B24","first-page":"4768","article-title":"\u201cA unified approach to interpreting model predictions,\u201d","volume-title":"Proceedings of the International Conference on Neural Information Processing Systems","author":"Lundberg","year":"2017"},{"key":"B25","doi-asserted-by":"publisher","first-page":"1","DOI":"10.1016\/j.artint.2018.07.007","article-title":"Explanation in artificial intelligence: insights from the social sciences","volume":"267","author":"Miller","year":"2019","journal-title":"Artif. Intell"},{"key":"B26","doi-asserted-by":"crossref","first-page":"652","DOI":"10.1145\/3461702.3462597","article-title":"\u201cTowards unifying feature attribution and counterfactual explanations: different means to the same end,\u201d","volume-title":"Proceedings of the AAAI\/ACM Conference on AI, Ethics, and Society","author":"Mothilal","year":"2021"},{"key":"B27","first-page":"4167","article-title":"\u201cExplaining visual models by causal attribution,\u201d","volume-title":"Proceedings of the International Conference on Computer Vision Workshop","author":"Parafita","year":"2019"},{"key":"B28","first-page":"570","article-title":"\u201cOn the impact of neighbourhood sampling to satisfy sufficiency and necessity criteria in explainable AI,\u201d","volume-title":"Proceedings of the Conference on Causal Learning and Reasoning","author":"Pawar","year":"2024"},{"key":"B29","doi-asserted-by":"crossref","DOI":"10.1017\/CBO9780511803161","author":"Pearl","year":"2009","journal-title":"Causality"},{"key":"B30","doi-asserted-by":"crossref","DOI":"10.1145\/3774643","article-title":"\u201cConcept-based explainable artificial intelligence: a survey,\u201d","volume-title":"ACM Computing Surveys","author":"Poeta","year":"2025"},{"key":"B31","author":"Ribeiro","year":"2016","journal-title":"Proceedings of the ACM SIGKDD International Conference on Knowledge Discovery and Data Mining"},{"key":"B32","first-page":"618","article-title":"\u201cGrad-CAM: visual explanations from deep networks via gradient-based localization,\u201d","volume-title":"Proceedings of the IEEE International Conference on Computer Vision","author":"Selvaraju","year":"2017"},{"key":"B33","doi-asserted-by":"publisher","first-page":"841","DOI":"10.2139\/ssrn.3063289","article-title":"Counterfactual explanations without opening the black box: Automated decisions and the GDPR","volume":"31","author":"Wachter","year":"2018","journal-title":"Harv. J. Law Technol"},{"key":"B34","doi-asserted-by":"publisher","first-page":"109111","DOI":"10.1016\/j.ijar.2023.109111","article-title":"Efficient computation of counterfactual bounds","volume":"171","author":"Zaffalon","year":"2024","journal-title":"Int. J. Approx. Reason"}],"container-title":["Frontiers in Artificial Intelligence"],"original-title":[],"link":[{"URL":"https:\/\/www.frontiersin.org\/articles\/10.3389\/frai.2025.1759000\/full","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2026,2,11]],"date-time":"2026-02-11T11:22:32Z","timestamp":1770808952000},"score":1,"resource":{"primary":{"URL":"https:\/\/www.frontiersin.org\/articles\/10.3389\/frai.2025.1759000\/full"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2026,2,11]]},"references-count":34,"alternative-id":["10.3389\/frai.2025.1759000"],"URL":"https:\/\/doi.org\/10.3389\/frai.2025.1759000","relation":{},"ISSN":["2624-8212"],"issn-type":[{"value":"2624-8212","type":"electronic"}],"subject":[],"published":{"date-parts":[[2026,2,11]]},"article-number":"1759000"}}