Multimodal LLMs Outperform Pathology Foundation Models in Cross-Domain Histological Similarity
cs.CV, cs.AI, cs.CL, cs.LG
Submitted: 2026-09-26
Updated: 2026-09-29
Code: https://github.com/bioptimus/releases
Terminology
Sources
- Current Pathology Foundation Models are unrobust to Medical Center Differences
- Multimodal Whole Slide Foundation Model for Pathology
- Phikon-v2, A large and public feature extractor for biomarker prediction
- Gemma 3 Technical Report
- Do Histopathological Foundation Models Eliminate Batch Effects? A Comparative Study
- Gemini Embedding: Generalizable Embeddings from Gemini
- A Multicentric Dataset for Training and Benchmarking Breast Cancer Segmentation in H&E Slides
- Hibou: A Family of Foundational Vision Transformers for Pathology
- DINOv2: Learning Robust Visual Features without Supervision
- OpenAI GPT-5 System Card
- Scanner-Induced Domain Shifts Undermine the Robustness of Pathology Foundation Models
- Virchow2: Scaling Self-Supervised Mixed Magnification Models in Pathology
- Training state-of-the-art pathology foundation models with orders of magnitude less data
- GLM-4.5V and GLM-4.1V-Thinking: Towards Versatile Multimodal Reasoning with Scalable Reinforcement Learning
Related papers
- Loss Knows Best: Detecting Annotation Errors in Videos via Loss Trajectories
- AnchorWeave: World-Consistent Video Generation with Retrieved Local Spatial Memories
- Benchmarking the Robustness of Foundation Models for Mammography under Domain Shift
- MambaX-Net: Dual-Input Mamba-Enhanced Cross-Attention Network for Longitudinal MRI Segmentation
- TeleOCR: Navigating Document Parsing Across Digital and Camera-Captured Documents
- A Survey on Efficient Vision-Language-Action Models