mlsanityData to 14 Oct 2026

Robustifying pathology foundation models via fine-tuning

arXiv 2607.22861 ↗First posted 24 Jul 2026cs.CV cs.AIcited by 0

Not on the map. Why?

Abstract (arXiv)

Pathology foundation models (FMs) produce powerful tile-level representations which remain sensitive to scanner and staining variability, undermining deployment across laboratories. We develop a novel fine-tuning recipe that improves the robustness of pathology FMs to acquisition factors. Applied to ten different FMs, our fine-tuning strategy consistently improves robustness for every model as well as downstream performance, with no observed trade-off. On average, it raises the PathoROB robustness index by 23% (from 0.72 to 0.87) and increases the overall cross-benchmark performance by 43% on Patho-Bench, HEST and THUNDER combined, with individual gains reaching up to 72% in robustness (Phikon-v2) and 76% in performance (Midnight-12k). We publicly release the fine-tuned versions of Phikon-v2 (Phaet) and Midnight-12k (Mascaret) at https://huggingface.co/wearewaiv/models.

Key facts. No citing papers in the map yet (data through 14 Oct 2026). Builds on 35 of its 63 references that are in the map. First posted on arXiv 24 Jul 2026.

Builds on

35 of 63 references.

References · 35

  1. 2023DINOv2: Learning Robust Visual Features without Supervisioncited by 7,974
  2. 2021Emerging Properties in Self-Supervised Vision Transformerscited by 6,968
  3. 2015Domain-Adversarial Training of Neural Networkscited by 3,044
  4. 2020Improved Baselines with Momentum Contrastive Learningcited by 2,287
  5. 2021Barlow Twins: Self-Supervised Learning via Redundancy Reductioncited by 1,838
  6. 2025DINOv3cited by 1,414
  7. 2018Attention-based Deep Multiple Instance Learningcited by 1,095
  8. 2021iBOT: Image BERT Pre-Training with Online Tokenizercited by 903
  9. 2021TransMIL: Transformer based Correlated Multiple Instance Learning for Whole Slide Image Classificationcited by 555
  10. 2019Quantifying the effects of data augmentation and stain color normalization in convolutional neural networks for computational pathologycited by 173
  11. 2024Multimodal Whole Slide Foundation Model for Pathologycited by 162
  12. 2024Virchow2: Scaling Self-Supervised Mixed Magnification Models in Pathologycited by 161
23 more
  1. 2020Self supervised contrastive learning for digital histopathologycited by 146
  2. 2024HEST-1k: A Dataset for Spatial Transcriptomics and Histology Image Analysiscited by 107
  3. 2024Phikon-v2, A large and public feature extractor for biomarker predictioncited by 74
  4. 2025Accelerating Data Processing and Benchmarking of AI Models for Pathologycited by 65
  5. 2020Self-Supervision Closes the Gap Between Weak and Strong Supervision in Histologycited by 43
  6. 2025Current Pathology Foundation Models are unrobust to Medical Center Differencescited by 39
  7. 2025Distilling foundation models for robust and efficient models in digital pathologycited by 36
  8. 2023Domain Generalization in Computational Pathology: Survey and Guidelinescited by 29
  9. 2021Self supervised learning improves dMMR/MSI detection from histology slides across multiple cancerscited by 23
  10. 2025Training state-of-the-art pathology foundation models with orders of magnitude less datacited by 21
  11. 2024Evaluating Computational Pathology Foundation Models for Prostate Cancer Grading under Distribution Shiftscited by 9
  12. 2025THUNDER: Tile-level Histopathology image UNDERstanding benchmarkcited by 9
  13. 2026Scanner-Induced Domain Shifts Undermine the Robustness of Pathology Foundation Modelscited by 9
  14. 2023ContriMix: Scalable stain color augmentation for domain generalization without domain labels in digital pathologycited by 7
  15. 2025MeDi: Metadata-Guided Diffusion Models for Mitigating Biases in Tumor Classificationcited by 7
  16. 2025Pathology Foundation Models are Scanner Sensitive: Benchmark and Mitigation with Contrastive ScanGen Losscited by 7
  17. 2026Atlas 2 -- Foundation models for clinical deploymentcited by 7
  18. 2025Beyond the Failures: Rethinking Foundation Models in Pathologycited by 6
  19. 2025Pathryoshka: Compressing Pathology Foundation Models via Multi-Teacher Knowledge Distillation with Nested Embeddingscited by 6
  20. 2025Beyond Diagnostic Performance: Revealing and Quantifying Ethical Risks in Pathology Foundation Modelscited by 4
  21. 2025SCORPION: Addressing Scanner-Induced Variability in Histopathologycited by 3
  22. 2026Enabling clinical use of foundation models for computational pathologycited by 1
  23. 2026A protocol for evaluating robustness to H&E staining variation in computational pathology modelscited by 1

Built on by

Not cited yet.