# Beyond the Failures: Rethinking Foundation Models in Pathology

arXiv: https://arxiv.org/abs/2510.23807 · HTML: https://mlsanity.com/paper/2510.23807/

## Key facts

- Cited by 6 papers in the map (data through 14 Oct 2026).
- Builds on 12 of its 20 references that are in the map.
- First posted on arXiv 27 Oct 2025.

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

- [On the Opportunities and Risks of Foundation Models](https://mlsanity.com/paper/2108.07258/) (2021)
- [Multimodal Whole Slide Foundation Model for Pathology](https://mlsanity.com/paper/2411.19666/) (2024)
- [On the Theoretical Limitations of Embedding-Based Retrieval](https://mlsanity.com/paper/2508.21038/) (2025)
- [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)
- [A Survey on Computational Pathology Foundation Models: Datasets, Adaptation Strategies, and Evaluation Tasks](https://mlsanity.com/paper/2501.15724/) (2025)
- [A Survey of Pathology Foundation Model: Progress and Future Directions](https://mlsanity.com/paper/2504.04045/) (2025)
- [Rotation-Agnostic Image Representation Learning for Digital Pathology](https://mlsanity.com/paper/2311.08359/) (2023)
- [Foundation Models -- A Panacea for Artificial Intelligence in Pathology?](https://mlsanity.com/paper/2502.21264/) (2025)
- [Learning Binary and Sparse Permutation-Invariant Representations for Fast and Memory Efficient Whole Slide Image Search](https://mlsanity.com/paper/2208.13653/) (2022)
- [Are the Latent Representations of Foundation Models for Pathology Invariant to Rotation?](https://mlsanity.com/paper/2412.11938/) (2024)
- [Universal and Transferable Attacks on Pathology Foundation Models](https://mlsanity.com/paper/2510.16660/) (2025)

## Built on by (most cited first, 6 of 6)

- [Do Pathology Foundation Models Encode Disease Progression? A Pseudotime Analysis of Visual Representations](https://mlsanity.com/paper/2601.21334/) (2026)
- [Epistemic Compression: The Case for Deliberate Ignorance in High-Stakes AI](https://mlsanity.com/paper/2603.25033/) (2026)
- [SIMPLER: H&E-Informed Representation Learning for Structured Illumination Microscopy](https://mlsanity.com/paper/2604.10334/) (2026)
- [Validation of Whole-Slide Foundation Models for Image Retrieval in TCGA Data](https://mlsanity.com/paper/2605.00902/) (2026)
- [MorphologyFM: A Foundation Model for Morphology-Aware Representation Learning from ECG and Pulse Oximetry Waveforms](https://mlsanity.com/paper/2607.09749/) (2026)
- [Robustifying pathology foundation models via fine-tuning](https://mlsanity.com/paper/2607.22861/) (2026)

## Abstract (arXiv)

Despite their successes in vision and language, foundation models have stumbled in pathology, revealing low accuracy, instability, and heavy computational demands. These shortcomings stem not from tuning problems but from deeper conceptual mismatches: dense embeddings cannot represent the combinatorial richness of tissue, and current architectures inherit flaws in self-supervision, patch design, and noise-fragile pretraining. Biological complexity and limited domain innovation further widen the gap. The evidence is clear-pathology requires models explicitly designed for biological images rather than adaptations of large-scale natural-image methods whose assumptions do not hold for tissue.