Article

MINA: linear probes reveal coding-sequence family signal in frozen DNA encoders up to protein composition

2026-05-28

Abstract excerpt

Frozen DNA encoders are widely used as feature extractors, but fine-tuning hides what their fixed embeddings already make linearly accessible. We introduce MINA (Model Interrogation of Nucleotide Architectures), a lightweight benchmark testing whether frozen DNA embeddings can recover (i) a 5-way protein-family label and (ii) the 1,536-d GenePT natural-language embedding for each gene, comparing coding-sequence (C...

Topics

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

Identifiers and source

Literature Corpus work: 8d8e5fdf-d758-5df5-a67d-9e08791ff5b7
DOI: 10.64898/2026.05.25.727711

Open publication

Related research

Semantic proximity does not establish scientific evidence.

Click a neighbor to travelStep 1 · 12 closest

MINA: linear probes reveal coding-sequence family signal in frozen DNA encoders up to protein compositionDOI 10.64898/2026.05.25.727711Read

Findings from Sparse Autoencoders for DNA Sequence Models: Motif Detectors, Reading-Frame Features, and the Scarcity of Regulatory LogicSemantic proximity Read

NucEL: Single-Nucleotide ELECTRA-Style Genomic Pre-training for Efficient and Interpretable RepresentationsSemantic proximity Read

Genomic Touchstone: Benchmarking Genomic Language Models in the Context of the Central DogmaSemantic proximity Read

Benchmarking DNA Foundation Models for Genomic and Genetic TasksSemantic proximity Read

DNABERT: pre-trained Bidirectional Encoder Representations from Transformers model for DNA-language in genomeSemantic proximity Read

Scaling Dense Representations for Single Cell with Transcriptome-Scale ContextSemantic proximity Read

Semantic fragment representations for coordinate-free analysis of genomics dataSemantic proximity Read

Genos-m: a foundation model for human-associated microbial genomesSemantic proximity Read

Genome modeling and design across all domains of life with Evo 2Semantic proximity Read

Mixture of Experts Enable Efficient and Effective Protein Understanding and DesignSemantic proximity Read

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

Fine-Tuning Transformers For Genomic TasksSemantic proximity Read