发表机构
Max Planck Institute for Informatics; Technical University of Darmstadt（马克斯·普朗克信息学研究所; 达姆施塔特工业大学）
机构由 AI 辅助整理，请以论文原文为准。AI 中文总结
AnyBottle提出一种构建紧凑任务特定概念瓶颈模型的方法，利用教师指导选择关键概念，实现更少概念且保持性能，证明无标注无需大型词汇表。
AI 中文摘要
概念瓶颈模型（CBMs）通过将预测路由到人类可解释的概念上，使得预测可检查且可干预，但最初需要概念标注。无标注变体消除了这一需求，但通常使用大型概念词汇表，在训练和推理时都是静态的，产生的瓶颈比任何任务或预测所需的都大，且更难以检查。我们提出了AnyBottle，一种构建紧凑、任务特定CBMs的单一配方。AnyBottle仅假设一个冻结的主干网络和一个无监督概念池，例如稀疏自编码器。然后，在同一主干网络上训练的黑盒教师模型指导选择：每轮添加最能解释瓶颈当前失败的概念，候选限制在教师/学生不一致的区域。通过在此选择顺序上使用嵌套丢弃进行训练，最终瓶颈可以从任何概念前缀准确预测，因此推理时对早期自信的输入使用较少概念，对困难输入使用更多概念。由于没有阶段是模态特定的，新领域和任务只需更换主干网络和概念池。在六个视觉和两个文本数据集以及两种教师范式上，AnyBottle产生的瓶颈概念更少，概念一致性高于无标注基线，同时接近黑盒参考。总体而言，AnyBottle表明，无标注并不意味着必须大型化：一个小的、发现出的词汇表可以像更大的、固定的词汇表一样具有表现力。
英文摘要
Concept bottleneck models (CBMs) make predictions inspectable and intervenable by routing them through human-interpretable concepts, but originally required concept annotations. Annotation-free variants remove this requirement, but typically use large concept vocabularies, static at both training and inference, producing bottlenecks larger than any task or prediction needs and harder to inspect. We propose AnyBottle, a single recipe for building compact, task-specific CBMs. AnyBottle assumes only a frozen backbone and an unsupervised concept pool, such as a sparse autoencoder. A black-box teacher trained on the same backbone then guides selection: each round adds the concept that best explains the bottleneck's current failures, with candidates restricted to regions of teacher/student disagreement. Trained with nested dropout over this selection order, the final bottleneck predicts accurately from any concept prefix, so inference spends fewer concepts on inputs it is confident about early and more on hard ones. Since no stage is modality-specific, a new domain and task requires swapping only the backbone and concept pool. Across six vision and two text datasets and two teacher paradigms, AnyBottle yields bottlenecks with fewer concepts and higher concept consistency than annotation-free baselines, while staying close to the black-box reference. Overall, AnyBottle shows that going annotation-free need not mean going large: a small, discovered vocabulary can be as expressive as a much larger, fixed one.
