---
title: C2S-Scale-Gemma-2-2B
description: Cell2Sentence-Scale (C2S-Scale) model from the van Dijk lab (Yale) and Google, fine-tuned from Gemma 2 2B. A cell's single-cell RNA-seq profile is written as a sentence of gene names ranked by expression, so the language model can predict cell type and tissue, generate cells and predict the effect of perturbations.
tags:
- biology
- cell-sentence
- computational-biology
- gemma2
- genomics
- scRNAseq
- single-cell
- text-generation
- transcriptomics
license: cc-by-4.0
source_url: https://huggingface.co/vandijklab/C2S-Scale-Gemma-2-2B
---

# C2S-Scale-Gemma-2-2B

> **Mirrored metadata.** This card is reproduced on Atlas for discovery. The artifact itself is not hosted here — follow the source link for the files, and refer to the upstream licence for terms of use.

Cell2Sentence-Scale (C2S-Scale) model from the van Dijk lab (Yale) and Google, fine-tuned from Gemma 2 2B. A cell's single-cell RNA-seq profile is written as a sentence of gene names ranked by expression, so the language model can predict cell type and tissue, generate cells and predict the effect of perturbations.

Upstream identifier: `vandijklab/C2S-Scale-Gemma-2-2B`.

**Source:** https://huggingface.co/vandijklab/C2S-Scale-Gemma-2-2B

**Licence:** cc-by-4.0

_This model is part of the Atlas demo catalogue. See the repository history for provenance._