A preprint introduces scHFBiMamba, a compact bidirectional state-space autoencoder built on the Mamba architecture. The model is designed to work with single-nucleus RNA-sequencing (snRNA-seq) data, a technique that profiles gene expression in individual nuclei from tissue samples.

The study evaluates whether this approach improves joint classification of three labels at once: species, cell type, and disease state. The application is heart failure (HF), where snRNA-seq can reveal cellular changes associated with the condition.

Because the abstract only outlines the evaluation goal, no results or performance comparisons are yet available. The source is a single arXiv preprint, so the claims are preliminary and have not been peer-reviewed.