← Back to all news
arXiv Machine Learning By Siddharth Setlur, Djordje Mihajlovic, Darrick Lee

Interpreting Latent Protein Language Model Features with Geometric Annotations

Read the original on arXiv Machine Learning →

The paper introduces a scalable method to interpret sparse autoencoder (SAE) features in the ESM-2 protein language model by leveraging geometrically inspired features of the protein α‑carbon backbone. Across 8M layers of ESM-2, a false discovery rate–controlled analysis shows that local geometry is significantly associated with many SAE features, revealing substructure within known biological labels and enabling annotation of unannotated metagenomic proteins. Ablation experiments demonstrate that removing these geometric features shifts ESM-2’s predicted contact maps toward the descriptor, linking mechanistic interpretability with structural biology.

Machine-generated by The Flow from the publisher's headline and feed description — not written or checked by a human. The full article lives at arXiv Machine Learning.

Related stories

arXiv Machine Learning
Jun 29

PairSAE: Mechanistic Interpretability from Pair Representations in Protein Co-Folding

arXiv:2606. 27440v1 Announce Type: new Abstract: Foundation models for structural biology have achieved remarkable performance in predicting biomolecular structure and show promise for the design of proteins and small molecules.

By Giosue Migliorini, Aristofanis Rontogiannis, Grigori Guitchounts, Nicholas Franklin, Axel Elaldi, Olivia Viessmann
llmsragdiffusionsafety
More like this →
arXiv Machine Learning
Aug 27

Interpreting Protein Language Model Embeddings via Orthogonal Projection for Protein Fitness Prediction

The paper introduces a method that uses orthogonal projection to remove the influence of known biochemical features from protein language model (PLM) embeddings, allowing the authors to assess how much these features contribute to protein fitness predictions. By applying this technique to high‑order and interaction effects, they demonstrate that eliminating these interpretable features reduces downstream classifier performance, indicating that PLM embeddings encode patterns correlated with biochemical properties. The authors also show that these biochemical features explain a substantial portion of the variance in the classifier’s predictions, suggesting that PLM embeddings capture biologically relevant information.

By Paulo Yanez Sarmiento, Pia Francesca Rissom, Manuel Pfeuffer, Marco Simnacher, Jordan F. Safer, Sumaiya Iqbal, Henrike O. Heyne, Nadja Klein, Bernhard Y. Renard
llmsragbenchmarks
More like this →