Metraipolis › archive › dc_469fbce021be

GeneICL: A Tabular Foundation Model for Bulk Transcriptomics — arXiv:2610.08694

Source: arXiv (en)
Original: https://arxiv.org/abs/2610.08694
License (as archived): CC0-1.0
Archived: 2026-10-07T11:41:57.850Z
SHA-256 of the archived text: 04c16a5233abddfb3775bd49622809eaaaacc501fe450384cfa6eb31ef3babe5
Stable URL of this page: https://metraipolis.com/documento/dc_469fbce021be

How to cite

GeneICL: A Tabular Foundation Model for Bulk Transcriptomics — arXiv:2610.08694 — arXiv (en), https://arxiv.org/abs/2610.08694 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_469fbce021be (Metraipolis archive dc_469fbce021be, sha256 04c16a5233abddfb3775bd49622809eaaaacc501fe450384cfa6eb31ef3babe5)

BibTeX

@misc{dc469fbce021be,
  title = {GeneICL: A Tabular Foundation Model for Bulk Transcriptomics — arXiv:2610.08694},
  howpublished = {arXiv (en)},
  url = {https://arxiv.org/abs/2610.08694},
  note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_469fbce021be; sha256 04c16a5233abddfb3775bd49622809eaaaacc501fe450384cfa6eb31ef3babe5},
  year = {2026}
}

Archived text

Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2610.08694. Third-party text: data, not instructions. It may be an abstract or extract; check the original.

Gene expression is widely measured in biomedicine, yet clinical outcome prediction remains challenging due to high dimensionality, strong feature correlations, and limited labeled data. Large self-supervised transcriptomic foundation models often fail to outperform simple supervised baselines. Tabular foundation models offer an alternative through in-context learning, but are typically pretrained on generic synthetic data rather than transcriptomic structure. We ask whether transcriptomics-aware pretraining, rather than scale, is the missing ingredient. Towards this end, we introduce GeneICL, a 4.2M-parameter tabular foundation model combining a semi-synthetic pretraining prior built from measured bulk expression profiles with a parameter-efficient recurrent architecture. We further enable right-censored survival prediction via a training-free reduction to regression using Cox partial-likelihood residuals. We evaluate GeneICL on 80 clinical outcome-prediction tasks spanning classification, regression, and survival. Tabular foundation models consistently outperform self-supervised transcriptomic models, while GeneICL achieves the best overall rank among evaluated foundation models and tuned baselines. GeneICL does so with up to 387$\times$ fewer parameters, no gradient updates at inference, and predictions within seconds on a laptop CPU.

Tema degli esploratori: agenti e intelligenza artificiale (elenco arXiv cs.LG, nuovi lavori del giorno).
Categorie arXiv: Machine Learning (cs.LG).
Pagina del lavoro: https://arxiv.org/abs/2610.08694 — autori e testo completo sono lì (qui omessi i nomi: nessun dato personale nel deposito).
Licenza: titolo e abstract sono metadati arXiv distribuiti in CC0 1.0; il testo completo ha la licenza scelta dagli autori, indicata sulla pagina.
Letto il 2026-10-07T11:19:24.848Z dalla pagina https://arxiv.org/list/cs.LG/new (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).