# Comparing Computational Pathology Foundation Models using Representational Similarity Analysis

arXiv: https://arxiv.org/abs/2509.15482 · HTML: https://mlsanity.com/paper/2509.15482/

## Key facts

- No citing papers in the map yet (data through 14 Oct 2026).
- Builds on 12 of its 29 references that are in the map.
- First posted on arXiv 18 Sep 2025.

## Builds on (most cited first, 12 of 12)

- [Learning Transferable Visual Models From Natural Language Supervision](https://mlsanity.com/paper/2103.00020/) (2021)
- [Domain-Adversarial Training of Neural Networks](https://mlsanity.com/paper/1505.07818/) (2015)
- [Quilt-1M: One Million Image-Text Pairs for Histopathology](https://mlsanity.com/paper/2306.11207/) (2023)
- [Virchow2: Scaling Self-Supervised Mixed Magnification Models in Pathology](https://mlsanity.com/paper/2408.00738/) (2024)
- [Benchmarking foundation models as feature extractors for weakly-supervised computational pathology](https://mlsanity.com/paper/2408.15823/) (2024)
- [Current Pathology Foundation Models are unrobust to Medical Center Differences](https://mlsanity.com/paper/2501.18055/) (2025)
- [Foundation Models in Computational Pathology: A Review of Challenges, Opportunities, and Impact](https://mlsanity.com/paper/2502.08333/) (2025)
- [Knowledge-enhanced Pretraining for Vision-language Pathology Foundation Model on Cancer Diagnosis](https://mlsanity.com/paper/2412.13126/) (2024)
- [Do Histopathological Foundation Models Eliminate Batch Effects? A Comparative Study](https://mlsanity.com/paper/2411.05489/) (2024)
- [Benchmarking Pathology Feature Extractors for Whole Slide Image Classification](https://mlsanity.com/paper/2311.11772/) (2023)
- [THUNDER: Tile-level Histopathology image UNDERstanding benchmark](https://mlsanity.com/paper/2507.07860/) (2025)
- [AdvDINO: Domain-Adversarial Self-Supervised Representation Learning for Spatial Proteomics](https://mlsanity.com/paper/2508.04955/) (2025)

## Abstract (arXiv)

Foundation models are increasingly developed in computational pathology (CPath) given their promise in facilitating many downstream tasks. While recent studies have evaluated task performance across models, less is known about the structure and variability of their learned representations. Here, we systematically analyze the representational spaces of six CPath foundation models using techniques popularized in computational neuroscience. The models analyzed span vision-language contrastive learning (CONCH, PLIP, KEEP) and self-distillation (UNI (v2), Virchow (v2), Prov-GigaPath) approaches. Through representational similarity analysis using H&E image patches from TCGA, we find that UNI2 and Virchow2 have the most distinct representational structures, whereas Prov-Gigapath has the highest average similarity across models. Having the same training paradigm (vision-only vs. vision-language) did not guarantee higher representational similarity. The representations of all models showed a high slide-dependence, but relatively low disease-dependence. Stain normalization decreased slide-dependence for all models by a range of 5.5% (CONCH) to 20.5% (PLIP). In terms of intrinsic dimensionality, vision-language models demonstrated relatively compact representations, compared to the more distributed representations of vision-only models. These findings highlight opportunities to improve robustness to slide-specific features, inform model ensembling strategies, and provide insights into how training paradigms shape model representations. Our framework is extendable across medical imaging domains, where probing the internal representations of foundation models can support their effective development and deployment.