arXiv cs.AI· Dibaloke Chanda, Praveen Madiraju·· 14 小时前AI 评分23
MOSAIC：迈向计算病理基础模型的通用潜在空间
MOSAIC: Towards a Universal Latent Space for Computational Pathology Foundation Models
AI 导读
MOSAIC 系统研究了 12 个 patch 级和 6 个 slide 级病理基础模型（PFM）的潜在空间相似性、对齐与互操作性，并评估多种对齐策略能否将独立学习的表征映射到共同潜在空间。研究通过跨模型一致性、检索、几何保持及分子与形态预测任务上的下游多实例学习进行评测，探索异构 PFM 能否从表征相似走向共享且可功能迁移的病理表征空间。代码与基准已公开，成果被 WACV 2027 接收。
正文 · 原文
Abstract:Pathology foundation models (PFMs) differ substantially in architecture, pretraining objective, training data and representation dimensionality, yet it remains unclear whether they learn fundamentally different feature spaces or alternative coordinate systems of a shared underlying representation. We present MOSAIC, a systematic study of latent-space similarity, alignment and interoperability across $12$ patch-level and $6$ slide-level PFMs. We first quantify representational agreement using complementary similarity measures and analyze how architecture, pretraining objective, network depth and magnification influence latent geometry. We then evaluate multiple alignment strategies for mapping independently learned representations into a common latent space. The resulting representations are evaluated through cross-model agreement, retrieval, geometry preservation, and downstream multiple-instance learning across molecular and morphological prediction tasks. MOSAIC provides a unified framework for studying whether heterogeneous PFMs can move beyond representational similarity toward a shared and functionally transferable pathology representation space. Our code, benchmarks and additional results are available at this https URL.
| Comments: | Accepted at WACV 2027 |
| Subjects: | Computer Vision and Pattern Recognition (cs.CV); Artificial Intelligence (cs.AI) |
| Cite as: | arXiv:2610.33059 [cs.CV] |
| (or arXiv:2610.33059v1 [cs.CV] for this version) | |
| https://doi.org/10.48550/arXiv.2610.33059 arXiv-issued DOI via DataCite |
Submission history
From: Dibaloke Chanda [view email]
[v1]
Mon, 19 Oct 2026 10:19:47 UTC (3,041 KB)
来源：arXiv cs.AI · arxiv.org