{"entity":{"id":"foundation-model","kind":"term","name":"Foundation model","aka":["foundation model","foundation models","FM","bulk-native foundation model","single-cell foundation model","large x model"],"tldr":"A foundation model is a large model trained on a vast amount of data without a specific task in mind, then adapted to many downstream uses.","summary":"In artificial intelligence a foundation model is a machine-learning model trained on vast datasets so that it can be applied across a wide range of use cases; large language models are the common example (Wikipedia). In oncology the term covers pathology encoders trained on millions of tiles (UNI, Virchow), single-cell models trained on tens of millions of cells (UCE, GeneCompass), bulk transcriptome models (BulkFormer) and genomic sequence models (Evo 2). A model pretrained on single cells is out of distribution for bulk tumour data, so bulk-native and single-cell-native are different claims.","asOf":"2026-09-24","wikipedia":"https://en.wikipedia.org/wiki/Foundation_model","links":[{"label":"Wikipedia","url":"https://en.wikipedia.org/wiki/Foundation_model"}],"tags":["cansim-terms"],"related":["idea-multimodal-foundation-model","cancer-ai-vocabulary"],"cancers":[],"sections":[],"technologies":["pathology-foundation-model","ai-compute-platforms"],"targets":[],"drugs":[],"companies":[],"institutions":[],"pathways":[],"terms":["self-supervised-pretraining","fine-tuning-vs-frozen","pathology-foundation-models"],"trials":[],"people":[],"bottlenecks":[],"keyPapers":[],"journals":[],"dependsOn":[],"notes":["Listed in the CanSim terms map 1.0.0 (docs/onco/terms.json, generated 2026-09-24), CC BY 4.0, attribution: CanSim project, an open, public-data-first cancer foundation-model programme; CanSim page path /terms/foundation-model."],"provenance":{"editedBy":"OnCo CanSim terms wave (Wikipedia summaries, standards and project pages, GDC and FDA pages, Europe PMC)","editedOn":"2026-09-24","note":"CanSim terms map 1.0.0 (docs/onco/terms.json, generated 2026-09-24), CC BY 4.0, attribution: CanSim project, an open, public-data-first cancer foundation-model programme"},"category":"Methods and models"},"route":"/terms/foundation-model/","neighbours":{"idea":[{"id":"idea-multimodal-foundation-model","kind":"idea","name":"Patient-level multimodal foundation models for treatment selection","route":"/ideas/idea-multimodal-foundation-model/"}],"term":[{"id":"cancer-ai-vocabulary","kind":"term","name":"Cancer AI vocabulary (CanSim terms map)","route":"/terms/cancer-ai-vocabulary/"},{"id":"fine-tuning-vs-frozen","kind":"term","name":"Fine-tuning versus frozen features, and LoRA","route":"/terms/fine-tuning-vs-frozen/"},{"id":"mixed-precision-gpu","kind":"term","name":"GPU training and mixed precision","route":"/terms/mixed-precision-gpu/"},{"id":"pathology-foundation-models","kind":"term","name":"Pathology foundation models: UNI, UNI2, Virchow2, CTransPath, CONCH, TITAN","route":"/terms/pathology-foundation-models/"},{"id":"self-supervised-pretraining","kind":"term","name":"Self-supervised pretraining (SSL)","route":"/terms/self-supervised-pretraining/"},{"id":"single-cell-foundation-models","kind":"term","name":"Single-cell and transcriptome foundation models: UCE, GeneCompass, BulkFormer, BulkRNABert","route":"/terms/single-cell-foundation-models/"},{"id":"single-cell-rna-seq","kind":"term","name":"Single-cell RNA sequencing (scRNA-seq, 10x Chromium)","route":"/terms/single-cell-rna-seq/"},{"id":"transformer-architecture","kind":"term","name":"Transformer and attention","route":"/terms/transformer-architecture/"},{"id":"virtual-cell-models","kind":"term","name":"Virtual cell models and in-silico perturbation screens","route":"/terms/virtual-cell-models/"}],"technology":[{"id":"ai-compute-platforms","kind":"technology","name":"AI compute and model platforms for oncology","route":"/technologies/ai-compute-platforms/"},{"id":"pathology-foundation-model","kind":"technology","name":"Pathology & radiology foundation models","route":"/technologies/pathology-foundation-model/"}]}}