# Interpreting biomedical and science models

Area: [Mechanistic interpretability](https://mlsanity.com/mechinterp/index.md). HTML: https://mlsanity.com/mechinterp/interpreting-biomedical-and-science-models/

## Key facts

- Interpreting biomedical and science models: 60 papers on arXiv since 2024 (data through 14 Oct 2026).
- 17 of its 60 mapped papers (28%) cite “InterPLM: Discovering Interpretable Features in Protein Language Models via Sparse Autoencoders” (2024, arXiv 2412.12101).
- 9 of its 60 mapped papers (15%) cite “Interpreting and Steering Protein Language Models through Sparse Autoencoders” (2025, arXiv 2502.09135).
- 1 new papers 24 Sep–14 Oct.

## Most cited here

- [InterPLM: Discovering Interpretable Features in Protein Language Models via Sparse Autoencoders](https://mlsanity.com/paper/2412.12101/) (2024), cited by 17 papers here
- [Interpreting and Steering Protein Language Models through Sparse Autoencoders](https://mlsanity.com/paper/2502.09135/) (2025), cited by 9 papers here
- [Towards Interpretable Protein Structure Prediction with Sparse Autoencoders](https://mlsanity.com/paper/2503.08764/) (2025), cited by 8 papers here
- [Learning biologically relevant features in a pathology foundation model using sparse autoencoders](https://mlsanity.com/paper/2407.10785/) (2024), cited by 7 papers here
- [An X-Ray Is Worth 15 Features: Sparse Autoencoders for Interpretable Radiology Report Generation](https://mlsanity.com/paper/2410.03334/) (2024), cited by 7 papers here
- [CytoSAE: Interpretable Cell Embeddings for Hematology](https://mlsanity.com/paper/2507.12464/) (2025), cited by 3 papers here
- [Systematic Evaluation of Single-Cell Foundation Model Interpretability Reveals Attention Captures Co-Expression Rather Than Unique Regulatory Signal](https://mlsanity.com/paper/2602.17532/) (2026), cited by 3 papers here
- [Sparse Autoencoders for Low-N Protein Function Prediction and Design](https://mlsanity.com/paper/2508.18567/) (2025), cited by 2 papers here
- [MedSAE: Dissecting MedCLIP Representations with Sparse Autoencoders](https://mlsanity.com/paper/2510.26411/) (2025), cited by 2 papers here
- [DILA: Dictionary Label Attention for Mechanistic Interpretability in High-dimensional Multi-label Medical Coding Prediction](https://mlsanity.com/paper/2409.10504/) (2024), cited by 1 papers here

## New (24 Sep–14 Oct)

- [Generative modeling of intrinsically disordered protein regions by reinforcing sparse autoencoder features](https://mlsanity.com/paper/2610.02189/) (2026-10-01)