{"about":{"site":"https://codewithpapers.app","non_affiliation":"Code with Papers and Syntology are not affiliated with, endorsed by, or sponsored by Papers with Code, Meta, or the pwc-archive mirror.","licence":"CC BY-SA 4.0","licence_url":"https://creativecommons.org/licenses/by-sa/4.0/legalcode","attribution":"https://codewithpapers.app/attribution","modified":"archive material modified by Syntology; see the attribution page"},"url":"/paper/making-sense-of-dependence-efficient-black","title":"Making Sense of Dependence: Efficient Black-box Explanations Using Dependence Measure","arxiv_id":"2206.06219","date":"2022-06-13","proceeding":null,"authors":["Paul Novello","Thomas Fel","David Vigouroux"],"abstract":"This paper presents a new efficient black-box attribution method based on Hilbert-Schmidt Independence Criterion (HSIC), a dependence measure based on Reproducing Kernel Hilbert Spaces (RKHS). HSIC measures the dependence between regions of an input image and the output of a model based on kernel embeddings of distributions. It thus provides explanations enriched by RKHS representation capabilities. HSIC can be estimated very efficiently, significantly reducing the computational cost compared to other black-box attribution methods. Our experiments show that HSIC is up to 8 times faster than the previous best black-box attribution methods while being as faithful. Indeed, we improve or match the state-of-the-art of both black-box and white-box attribution methods for several fidelity metrics on Imagenet with various recent model architectures. Importantly, we show that these advances can be transposed to efficiently and faithfully explain object detection models such as YOLOv4. Finally, we extend the traditional attribution methods by proposing a new kernel enabling an ANOVA-like orthogonal decomposition of importance scores based on HSIC, allowing us to evaluate not only the importance of each image patch but also the importance of their pairwise interactions. Our implementation is available at https://github.com/paulnovello/HSIC-Attribution-Method.","url_abs":"https://arxiv.org/abs/2206.06219v3","url_pdf":"https://arxiv.org/pdf/2206.06219v3.pdf","source":{"archive":"pwc-archive (Hugging Face), CC BY-SA 4.0","snapshot":"2025-07-28","licence_url":"https://creativecommons.org/licenses/by-sa/4.0/legalcode","row_kind":"abstracts"},"code_links":[{"paper_slug":"making-sense-of-dependence-efficient-black","repo_url":"https://github.com/paulnovello/hsic-attribution-method","is_official":1,"mentioned_in_paper":1,"mentioned_in_github":0,"framework":"tf","reach":{"status":"ok"}}],"tasks":[{"task_slug":"error-understanding","task_name":"Error Understanding"},{"task_slug":"image-attribution","task_name":"Image Attribution"},{"task_slug":"interpretability-techniques-for-deep-learning","task_name":"Interpretability Techniques for Deep Learning"},{"task_slug":"object-detection","task_name":"Object Detection"},{"task_slug":"object-detection-1","task_name":"object-detection"}],"methods":[{"method_slug":"1x1-convolution","method_name":"1x1 Convolution"},{"method_slug":"average-pooling","method_name":"Average Pooling"},{"method_slug":"batch-normalization","method_name":"Batch Normalization"},{"method_slug":"bottom-up-path-augmentation","method_name":"Bottom-up Path Augmentation"},{"method_slug":"cspdarknet53","method_name":"CSPDarknet53"},{"method_slug":"convolution","method_name":"Convolution"},{"method_slug":"cosine-annealing","method_name":"Cosine Annealing"},{"method_slug":"cutmix","method_name":"CutMix"},{"method_slug":"dropblock","method_name":"DropBlock"},{"method_slug":"fpn","method_name":"FPN"},{"method_slug":"global-average-pooling","method_name":"Global Average Pooling"},{"method_slug":"grid-sensitive","method_name":"Grid Sensitive"},{"method_slug":"label-smoothing","method_name":"Label Smoothing"},{"method_slug":"logistic-regression","method_name":"Logistic Regression"},{"method_slug":"max-pooling","method_name":"Max Pooling"},{"method_slug":"pafpn","method_name":"PAFPN"},{"method_slug":"relu","method_name":"ReLU"},{"method_slug":"residual-connection","method_name":"Residual Connection"},{"method_slug":"sigmoid-activation","method_name":"Sigmoid Activation"},{"method_slug":"softmax","method_name":"Softmax"},{"method_slug":"spatial-pyramid-pooling","method_name":"Spatial Pyramid Pooling"},{"method_slug":"tanh-activation","method_name":"Tanh Activation"},{"method_slug":"yolov3","method_name":"YOLOv3"},{"method_slug":"yolov4","method_name":"YOLOv4"},{"method_slug":"k-means-clustering","method_name":"k-Means Clustering"}],"datasets_introduced":[],"methods_introduced":[],"results":[{"leaderboard":"/sota/error-understanding-on-cub-200-2011-1","task":"Error Understanding","dataset":"CUB-200-2011","model":"HSIC-Attribution","rank_in_archive_order":4,"of":4,"metrics":{"Average highest confidence (EfficientNetV2-M)":"0.2679","Average highest confidence (MobileNetV2)":"0.2914","Average highest confidence (ResNet-101)":"0.2493","Insertion AUC score (EfficientNetV2-M)":"0.1611","Insertion AUC score (MobileNetV2)":"0.1635","Insertion AUC score (ResNet-101)":"0.1446"},"uses_additional_data":false},{"leaderboard":"/sota/error-understanding-on-cub-200-2011-resnet","task":"Error Understanding","dataset":"CUB-200-2011 (ResNet-101)","model":"HSIC-Attribution","rank_in_archive_order":3,"of":3,"metrics":{"Average highest confidence":"0.2493","Insertion AUC score":"0.1446"},"uses_additional_data":false},{"leaderboard":"/sota/image-attribution-on-cub-200-2011-1","task":"Image Attribution","dataset":"CUB-200-2011","model":"HSIC-Attribution","rank_in_archive_order":4,"of":8,"metrics":{"Deletion AUC score (ResNet-101)":"0.0647","Insertion AUC score (ResNet-101)":"0.6843"},"uses_additional_data":false},{"leaderboard":"/sota/image-attribution-on-celeba","task":"Image Attribution","dataset":"CelebA","model":"HSIC-Attribution","rank_in_archive_order":3,"of":8,"metrics":{"Deletion AUC score (ArcFace ResNet-101)":"0.1151","Insertion AUC score (ArcFace ResNet-101)":"0.5692"},"uses_additional_data":false},{"leaderboard":"/sota/image-attribution-on-vggface2","task":"Image Attribution","dataset":"VGGFace2","model":"HSIC-Attribution","rank_in_archive_order":2,"of":8,"metrics":{"Deletion AUC score (ArcFace ResNet-101)":"0.1317","Insertion AUC score (ArcFace ResNet-101)":"0.6694"},"uses_additional_data":false},{"leaderboard":"/sota/interpretability-techniques-for-deep-learning-1","task":"Interpretability Techniques for Deep Learning","dataset":"CelebA","model":"HSIC-Attribution","rank_in_archive_order":2,"of":7,"metrics":{"Insertion AUC score":"0.5692"},"uses_additional_data":false}],"syntology":{"syntology_url":null,"atlas_url":"https://app.syntology.ai/?focus=2206.06219","mcp":{"get_harvested_code_for_paper":{"arxiv_id":"2206.06219"}},"developers":"https://syntology.ai/developers","read_at":"2026-09-24T18:15:14+00:00","read_at_is":"when the build read Syntology's graph, not when any sample ran","claim":"Per-sample execution status on synthesized fixtures; not a correctness claim about the paper. Samples come from repositories linked to the paper, official or community; repo_kind says which.","repos":[{"provenance":"deterministic:regex_extraction","url":"https://github.com/paulnovello/HSIC-Attribution-Method","reach":{"status":"ok"}},{"provenance":"external:paperswithcode_snapshot_2025-07-28","url":"https://github.com/paulnovello/hsic-attribution-method","reach":{"status":"ok"}}],"summary":{"ran_draft_wrong":1},"by_repo_kind":{"official":{"samples":1,"ran":1,"repositories":1}},"repo_kind_vocabulary":{"official":"The archive marks this repository official for the paper","named_in_paper":"The archive records that the paper mentions this repository; it is not marked official","listed":"In the archive's code links for this paper, not marked official and not recorded as mentioned in the paper","found_in_text":"Syntology found this repository in the paper's own text; whether it is the authors' implementation is not asserted","community":"Not in the archive's code links for this paper; a community repository Syntology harvested"},"n_pointer_only_for_licence":1,"samples":[{"code_sha256_prefix":"8bce65a80b5ee368","entry":"get_alpha_cmap","repo":"paulnovello/hsic-attribution-method","repo_kind":"official","path":"utils.py","file_url":"https://github.com/paulnovello/hsic-attribution-method/blob/HEAD/utils.py","link_basis":"first_harvest_node","language":"python","status":"ran_draft_wrong","verification_level":1,"contract_check":"OUTPUT_MISDECLARED","metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"NONE","inline_ok":false,"mcp_get_code":{"code_sha256":"8bce65a80b5ee368"}}]},"arxiv_metadata":null,"syntology_extracted_results":null}