# ProtSAE: Disentangling and Interpreting Protein Language Models via Semantically-Guided Sparse Autoencoders

arXiv: https://arxiv.org/abs/2509.05309 · HTML: https://mlsanity.com/paper/2509.05309/

## Key facts

- Cited by 3 papers in the map (data through 14 Oct 2026).
- Placed in Interpreting biomedical and science models, one of 13 directions in Mechanistic interpretability.
- Builds on 14 of its 38 references that are in the map.
- First posted on arXiv 26 Aug 2025.

## Placed in

- Mechanistic interpretability / [Interpreting biomedical and science models](https://mlsanity.com/mechinterp/interpreting-biomedical-and-science-models/)

## Builds on (most cited first, 14 of 14)

- [Interpretability in the Wild: a Circuit for Indirect Object Identification in GPT-2 small](https://mlsanity.com/paper/2211.00593/) (2022)
- [Finding Neurons in a Haystack: Case Studies with Sparse Probing](https://mlsanity.com/paper/2305.01610/) (2023)
- [Interpreting CLIP with Sparse Linear Concept Embeddings (SpLiCE)](https://mlsanity.com/paper/2402.10376/) (2024)
- [Transformer visualization via dictionary learning: contextualized embedding as a linear superposition of transformer factors](https://mlsanity.com/paper/2103.15949/) (2021)
- [BERTology Meets Biology: Interpreting Attention in Protein Language Models](https://mlsanity.com/paper/2006.15222/) (2020)
- [InterPLM: Discovering Interpretable Features in Protein Language Models via Sparse Autoencoders](https://mlsanity.com/paper/2412.12101/) (2024)
- [ProLLaMA: A Protein Large Language Model for Multi-Task Protein Language Processing](https://mlsanity.com/paper/2402.16445/) (2024)
- [EL Embeddings: Geometric construction of models for the Description Logic EL ++](https://mlsanity.com/paper/1902.10499/) (2019)
- [Codebook Features: Sparse and Discrete Interpretability for Neural Networks](https://mlsanity.com/paper/2310.17230/) (2023)
- [Sparse Autoencoder Features for Classifications and Transferability](https://mlsanity.com/paper/2502.11367/) (2025)
- [Interpreting and Steering LLMs with Mutual Information-based Explanations on Sparse Autoencoders](https://mlsanity.com/paper/2502.15576/) (2025)
- [Concept Bottleneck Language Models For protein design](https://mlsanity.com/paper/2411.06090/) (2024)
- [Sparse Shift Autoencoders for Identifying Concepts from Large Language Model Activations](https://mlsanity.com/paper/2502.12179/) (2025)
- [Controllable Protein Sequence Generation with LLM Preference Optimization](https://mlsanity.com/paper/2501.15007/) (2025)

## Built on by (most cited first, 3 of 3)

- [Induction Meets Biology: Mechanisms of Repeat Detection in Protein Language Models](https://mlsanity.com/paper/2602.23179/) (2026)
- [No Generation without Representation: Efficient Causal Protein Language Models Enable Zero-Shot Fitness Estimation](https://mlsanity.com/paper/2602.01845/) (2026)
- [Interpreting Latent Protein Language Model Features with Geometric Annotations](https://mlsanity.com/paper/2608.26419/) (2026)

## Abstract (arXiv)

Sparse Autoencoder (SAE) has emerged as a powerful tool for mechanistic interpretability of large language models. Recent works apply SAE to protein language models (PLMs), aiming to extract and analyze biologically meaningful features from their latent spaces. However, SAE suffers from semantic entanglement, where individual neurons often mix multiple nonlinear concepts, making it difficult to reliably interpret or manipulate model behaviors. In this paper, we propose a semantically-guided SAE, called ProtSAE. Unlike existing SAE which requires annotation datasets to filter and interpret activations, we guide semantic disentanglement during training using both annotation datasets and domain knowledge to mitigate the effects of entangled attributes. We design interpretability experiments showing that ProtSAE learns more biologically relevant and interpretable hidden features compared to previous methods. Performance analyses further demonstrate that ProtSAE maintains high reconstruction fidelity while achieving better results in interpretable probing. We also show the potential of ProtSAE in steering PLMs for downstream generation tasks.