Article

Detecting and quantifying overparametrization in RNA language models with REDIAL

2026-05-12

Abstract excerpt

<h4>ABSTRACT</h4> While RNA language models (LMs) have served as foundation models (FMs) to advanced structural prediction, their evaluation relies heavily on supervised downstream tasks. Such tasks can often mask FM inefficiencies and reflect downstream training set memorization. To address this, here we introduce REDIAL (RNA Embedding perturbation Diagnostics for Language models), a zero-shot, unsupervised fram...

Topics

Open a Topic to create a Post that cites this publication.

Identifiers and source

Literature Corpus work: 1e00149e-53ce-5749-a17a-2428155c427c
DOI: 10.64898/2026.05.11.724344

Open publication

Related research

Semantic proximity does not establish scientific evidence.

Click a neighbor to travelStep 1 · 12 closest

Detecting and quantifying overparametrization in RNA language models with REDIALDOI 10.64898/2026.05.11.724344Read

Fine-Tuning Protein Language Models Unlocks the Potential of Underrepresented Viral ProteomesSemantic proximity Read

Pretrained protein language models choose between sequence novelty and structural completenessSemantic proximity Read

Improving Variant Effect Prediction by Steering Sparse Mechanistic Features in Protein Language ModelsSemantic proximity Read

Capturing Natural Evolution in Function-guided RNA Design via Genomic Foundation ModelsSemantic proximity Read

Adapting protein language models for structure-conditioned designSemantic proximity Read

Safeguarding open-weight genomic foundation models through weight lockingSemantic proximity Read

A foundation language model to decipher diverse regulation of RNAsSemantic proximity Read

InstructPLM: Aligning Protein Language Models to Follow Protein Structure InstructionsSemantic proximity Read

InterPLM: Discovering Interpretable Features in Protein Language Models via Sparse AutoencodersSemantic proximity Read

Benchmarking Large Language Models for Predicting Therapeutic Antisense Oligonucleotide EfficacySemantic proximity Read

InstructPLM: Aligning Protein Language Models to Follow Protein Structure InstructionsSemantic proximity Read

LoMuS: Low-Rank Adaptation with Multimodal Representations Improves Protein Stability PredictionSemantic proximity Read