Reproduce “SurvICL: a Foundation Model for in-context Survival Analysis” on OpenResearch
Turn your favorite coding agents into research agents that review literature, develop hypotheses, and run experiments.
Universal app for Apple silicon and Intel. Requires macOS 11 or later.
About this paper
Julie Alberge, Linus Bleistein, Judith Abécassis, Marine Le Morvan, Gaël Varoquaux, David Holzmüller
SurvICL predicts survival curves from censored clinical data using an in-context model pretrained on synthetic survival examples, where censored context rows are paired with simulator-known event-time targets. Its proportional-odds prior allows hazard ratios to vary over time. Across 63 held-out datasets, it wins 82.9% of IBS comparisons versus Cox’s 71.3%; on datasets over 1,000 samples, SurvivalBoost leads in IBS.
Summary by alphaXiv.
Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.