{"ID":5675965,"CreatedAt":"2026-07-03T01:40:09.565152011Z","UpdatedAt":"2026-07-04T19:15:18.090787218Z","DeletedAt":null,"paper_url":"https://arxiv.org/abs/2607.01416","arxiv_id":"2607.01416","title":"Beyond Heatmaps: Unsupervised Concept-Graph Reasoning for Interpretable Visual Explanation","abstract":"Concept Bottleneck Models (CBMs) provide an intrinsically interpretable alternative to post-hoc explanations. However, existing CBMs often rely on predefined concept vocabularies or supervised annotations, lack explicit concept grounding, and summarize each concept with a single image-level score -- discarding spatial recurrence and inter-concept dependencies. We propose a Graph-based Concept Bottleneck Model (G-CBM), an intrinsically interpretable framework that performs unsupervised concept discovery via Non-negative Matrix Factorization (NMF) and represents the discovered concepts as nodes in a per-image concept-graph representation. G-CBM matches region-level features to these concept nodes -- providing concept grounding and capturing concept recurrence across the image -- and applies a \\emph{tunable concept filtering threshold} $τ$ to suppress weak region-level features. A Graph Attention Network (GAT) then performs concept-level reasoning by modeling nonlinear dependencies across nodes. Across ImageNet, HAM10000, PH2, and Derm7pt, G-CBM achieves an average relative AUC improvement of 3.7\\% over a ResNet-50 baseline. Concept filtering frequently improves predictive performance while inducing selective concept use, achieving peak AUC of $0.96$ on PH2 with only 2 of 10 concepts and 0.92 on HAM10000 with 3.8 of 9 concepts. On dermoscopy benchmarks, G-CBM is competitive with supervised approaches requiring external annotations. Deletion/insertion analyses with random ablation controls show that the learned concept ranking faithfully reflects model predictions.","short_abstract":"Concept Bottleneck Models (CBMs) provide an intrinsically interpretable alternative to post-hoc explanations. However, existing CBMs often rely on predefined concept vocabularies or supervised annotations, lack explicit concept grounding, and summarize each concept with a single image-level score -- discarding spatial...","url_abs":"https://arxiv.org/abs/2607.01416","url_pdf":"https://arxiv.org/pdf/2607.01416v1","authors":"[\"Md Mohasin Hossain\",\"Anar Amirli\",\"Robert Leist\",\"Md Abdul Kadir\",\"Daniel Sonntag\"]","published":"2026-07-01T19:21:39Z","proceeding":"cs.CV","tasks":"[\"cs.CV\"]","methods":"[]","has_code":false}