SIMPLER learns biologically grounded SIM representations by progressively aligning them with H&E images through multiple self-supervised objectives, outperforming scratch-trained or H&E-only models on downstream tasks like multiple instance learning and clustering.
Beyond the Failures: Rethinking Foundation Models in Pathology
2 Pith papers cite this work. Polarity classification is still indexing.
2
Pith papers citing it
abstract
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.
citation-role summary
background 1
citation-polarity summary
fields
cs.CV 2years
2026 2verdicts
UNVERDICTED 2roles
background 1polarities
background 1representative citing papers
Benchmarking on TCGA shows TITAN foundation model edges out others for whole-slide retrieval but with only ~68% average accuracy, high organ-to-organ variation, and no consistent winner over patch-level baselines.
citing papers explorer
-
SIMPLER: H&E-Informed Representation Learning for Structured Illumination Microscopy
SIMPLER learns biologically grounded SIM representations by progressively aligning them with H&E images through multiple self-supervised objectives, outperforming scratch-trained or H&E-only models on downstream tasks like multiple instance learning and clustering.
-
Validation of Whole-Slide Foundation Models for Image Retrieval in TCGA Data
Benchmarking on TCGA shows TITAN foundation model edges out others for whole-slide retrieval but with only ~68% average accuracy, high organ-to-organ variation, and no consistent winner over patch-level baselines.