# Layer Probing Improves Kinase Functional Prediction with Protein Language Models

arXiv: https://arxiv.org/abs/2512.00376 · HTML: https://mlsanity.com/paper/2512.00376/

## Key facts

- Cited by 1 papers in the map (data through 14 Oct 2026).
- Builds on 1 of its 9 references that are in the map.
- First posted on arXiv 29 Nov 2025.

## Builds on (most cited first, 1 of 1)

- [A Primer in BERTology: What we know about how BERT works](https://mlsanity.com/paper/2002.12327/) (2020)

## Built on by (most cited first, 1 of 1)

- [Task- and dataset-specific information in protein language models](https://mlsanity.com/paper/2608.12090/) (2026)

## Abstract (arXiv)

Protein language models (PLMs) have transformed sequence-based protein analysis, yet most applications rely only on final-layer embeddings, which may overlook biologically meaningful information encoded in earlier layers. We systematically evaluate all 33 layers of ESM-2 for kinase functional prediction using both unsupervised clustering and supervised classification. We show that mid-to-late transformer layers (layers 20-33) outperform the final layer by 32 percent in unsupervised Adjusted Rand Index and improve homology-aware supervised accuracy to 75.7 percent. Domain-level extraction, calibrated probability estimates, and a reproducible benchmarking pipeline further strengthen reliability. Our results demonstrate that transformer depth contains functionally distinct biological signals and that principled layer selection significantly improves kinase function prediction.