arXiv：cs.AI· Cl\'ement B\'enard, Manon Arfib, Christophe Labreuche, Victor Qu\'etu·· 20 小时前
Hoeffding 概念瓶颈模型（HCBM）：面向 overhead 图像的非线性稀疏概念聚合
Hoeffding Concept Bottleneck Models with Applications to Overhead Images
AI 导读
研究者提出 Hoeffding 概念瓶颈模型（HCBM），基于梯度提升树的 Hoeffding 函数分解，对概念分数做非线性、稀疏聚合，并用素蕴涵式生成紧凑预测，以解决现有 CBM 线性聚合概念数量多、削弱可解释性并易致信息泄漏的问题。HCBM 在理论上对概念间泄漏具有鲁棒性，实验表现优于稀疏线性 CBM，并可扩展至目标检测，在 overhead 图像这一高难度场景中展现出高性能。
正文
Abstract:Explainability of deep learning algorithms is critical for computer-vision applications with high-stakes decisions. Concept bottleneck models (CBM) have recently shown promising performance to provide explainable and accurate predictions for classification problems, based on a bottleneck of high-level concepts. Existing CBM methods rely on a linear aggregation of the concept scores to compute predictions. However, a large number of concepts is often used in this linear approach, which undermines explainability and favors information leakage. In general, the underlying relation between concepts and output logits is not linear. Therefore, we introduce Hoeffding Concept Bottleneck Models (HCBM), which build on the Hoeffding functional decomposition of gradient-boosted trees to provide non-linear and sparse aggregations of concept scores, and generate compact predictions using prime implicants. HCBM have strong theoretical properties of robustness to interconcept leakage, and outperform sparse linear CBM in practice, as shown in extensive experiments. Beyond classification, HCBM can be adapted to object detection, and we focus on a challenging case with overhead images to show the high performance of HCBM in these settings.
| Subjects: | Machine Learning (cs.LG); Artificial Intelligence (cs.AI); Machine Learning (stat.ML) |
| Cite as: | arXiv:2606.00082 [cs.LG] |
| (or arXiv:2606.00082v2 [cs.LG] for this version) | |
| https://doi.org/10.48550/arXiv.2606.00082 arXiv-issued DOI via DataCite |
Submission history
From: Clement Benard [view email] [via CCSD proxy]
[v1]
Fri, 22 May 2026 14:00:30 UTC (3,125 KB)
[v2]
Mon, 12 Oct 2026 09:53:16 UTC (3,128 KB)
来源：arXiv：cs.AI · arxiv.org