标题： CytoSAE：用于血液学的可解释细胞嵌入 显示英文标题

标题： CytoSAE: Interpretable Cell Embeddings for Hematology

摘要： 稀疏自编码器（SAEs）作为一种有前景的工具，用于基于变压器的基础模型的机制可解释性。 最近，SAEs也被应用于视觉领域，使得能够发现视觉概念并将它们逐补丁地归因于变压器模型中的标记。 虽然越来越多的基础模型被用于医学影像，但解释其推理的工具仍然缺乏。 在本工作中，我们展示了SAEs在血液学中的适用性。 我们提出了CytoSAE，一种在超过40,000张外周血单细胞图像上训练的稀疏自编码器。 CytoSAE可以推广到各种和域外的数据集，包括骨髓细胞学，在其中它识别出形态学相关的概念，我们通过医学专家进行了验证。 此外，我们展示了CytoSAE可以生成患者特异性和疾病特异性的概念的情况，从而能够在补丁级别检测特征性细胞和局部细胞异常。 我们量化了概念对患者级AML亚型分类任务的影响，并表明CytoSAE的概念在性能上与最先进方法相当，同时在亚细胞水平上提供可解释性。 源代码和模型权重可在https://github.com/dynamical-inference/cytosae获取。 显示英文摘要

摘要： Sparse autoencoders (SAEs) emerged as a promising tool for mechanistic interpretability of transformer-based foundation models. Very recently, SAEs were also adopted for the visual domain, enabling the discovery of visual concepts and their patch-wise attribution to tokens in the transformer model. While a growing number of foundation models emerged for medical imaging, tools for explaining their inferences are still lacking. In this work, we show the applicability of SAEs for hematology. We propose CytoSAE, a sparse autoencoder which is trained on over 40,000 peripheral blood single-cell images. CytoSAE generalizes to diverse and out-of-domain datasets, including bone marrow cytology, where it identifies morphologically relevant concepts which we validated with medical experts. Furthermore, we demonstrate scenarios in which CytoSAE can generate patient-specific and disease-specific concepts, enabling the detection of pathognomonic cells and localized cellular abnormalities at the patch level. We quantified the effect of concepts on a patient-level AML subtype classification task and show that CytoSAE concepts reach performance comparable to the state-of-the-art, while offering explainability on the sub-cellular level. Source code and model weights are available at https://github.com/dynamical-inference/cytosae.