Article

Rotation, rescaling, or replacement? Decomposing organ adaptation in single-cell transformer representations

2026-08-06

Abstract excerpt

<title>Abstract</title> <p>Single-cell transformers are usually studied one checkpoint at a time, which makes it impossible to separate what the architecture imposes on a representation from what the training corpus writes into it. The three released scGPT checkpoints offer a controlled alternative: architecture, objective, optimizer settings and gene vocabulary are identical, and only the pre-training corpus dif...

Topics

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

Identifiers and source

Literature Corpus work: ae99343b-4a49-5fac-a79c-6c29130e523a
DOI: 10.21203/rs.3.rs-10595347/v1

Open publication

Related research

Semantic proximity does not establish scientific evidence.

Click a neighbor to travelStep 1 · 12 closest

Rotation, rescaling, or replacement? Decomposing organ adaptation in single-cell transformer representationsDOI 10.21203/rs.3.rs-10595347/v1Read

Single-cell foundation models reveal context-sensitive cancer programmes under subtype shiftSemantic proximity Read

Gene-Family Encoding Boosts Domain-Adapted Single-Cell Language ModelsSemantic proximity Read

TRAILBLAZER: generative multicellular perturbation model of biologySemantic proximity Read

Hierarchical Interpretation of Out-of-Distribution Cells Using Bottlenecked TransformerSemantic proximity Read

Geometric Representations of Knowledge Inside Biological Large Language Models: an Empirical AnalysisSemantic proximity Read

Query to reference single-cell integration with transfer learningSemantic proximity Read

Do Geometric Outliers Identify Important Genes in Single-Cell Foundation Models?Semantic proximity Read

A controlled human–macaque basal ganglia transfer benchmark from HMBA expressionSemantic proximity Read

Tissue tearing degrades optimal-transport and diffeomorphic registration of spatial transcriptomics beyond displacement magnitude: a multi-seed deformation benchmark and a supervised graph cross-attention proof-of-conceptSemantic proximity Read

Human to Mouse Cross-Species Transfer Learning forElectrophysiology-to-Transcriptomics Mapping in CorticalGABAergic InterneuronsSemantic proximity Read

1 CodonBERT and ESM-2 Embedding Spaces Share an Evolutionarily Con-served Paired Geometry Encoding Synonymous Codon InformationSemantic proximity Read

Recursive Transport Topology: Diagnostics, Tunable Abstraction, and Latent Enhancement in Biological and World-Model RepresentationsSemantic proximity Read