Loading
Life sciences · Preprint
arXiv · September 4, 2026
Raises a question worth testing. It does not answer one.
This preprint describes a computational study comparing three adaptation strategies (zero-shot reformulation, classification fine-tuning, and survival-head adaptation) for applying tabular foundation models to censored time-to-event prediction across benchmark datasets. The work is methodological and exploratory, identifying that Cox proportional hazards provides the most stable performance metric (Integrated Brier Score) on larger datasets, while optimal strategy varies by data size and statistical structure, but does not validate these approaches in clinical or real-world prospective settings.
Computational benchmark evaluation across multiple datasets. Tabular benchmark datasets with time-to-event and censoring information; composition and clinical context not specified.. Intervention: Tabular foundation models with Cox, DeepHit, and MTLR adaptation interfaces for survival prediction.. Compared with: Zero-shot inference versus supervised fine-tuning versus survival-head adaptation; relative performance across data regimes..
Zero-shot inference effective on smaller single-risk datasets; supervised adaptation increasingly advantageous as datasets scale Cox provides most reliably strong interface, especially for Integrated Brier Score on larger datasets DeepHit relatively stronger for time-dependent Concordance Index than for Integrated Brier Score
Safety was not reported in the material analysed. Check the source before drawing any conclusion about harm.
The source did not state who this applies to in practice.
This is a methodological comparison study using computational modeling and benchmark datasets to explore adaptation strategies for tabular foundation models in survival analysis; it raises questions about optimal interfaces rather than testing a clinical hypothesis or intervention.
As stated by the source record.
Quoted from the source exactly as published.
Graded across the dimensions that decide whether you should act, each from what the source actually supports. There is no single score, and where a dimension was not assessed it says so.
Tabular foundation models (TabFMs) achieve strong performance on structured data, particularly for standard classification and regression problems. Yet, extending them to censored time-to-event prediction is challenging because it requires properly handling censoring and event-time dynamics. Building on our prior work, we further link TabFMs with CoxPH and DeepHit and revise the context-resampled training procedure. We evaluate temporal zero-shot reformulation, classification-based fine-tuning, and survival-head adaptation using frozen TabFM backbones on 74 single-risk data sets, and we additionally study 4 competing-risk data sets. Zero-shot inference is effective on smaller single-risk data sets, whereas supervised adaptation becomes increasingly advantageous as data sets scale. Cox provides the most reliably strong interface, especially for Integrated Brier Score (IBS) on larger data sets. DeepHit is relatively stronger for the time-dependent Concordance Index than for IBS, while cause-specific MTLR ranks highest among the TabFM survival heads in the four-data-set competing-risk analysis. Classification fine-tuning becomes more competitive with zero-shot inference as data sets grow but remains weaker for probabilistic prediction. Overall, our results indicate that effective TabFM transfer depends on the data regime and on the statistical structure represented by the chosen adaptation interface. The implementation scripts used for this work are available at https://github.com/kaylode/survival-fm.
Taken from the source record, never inferred. Follow any of these and new work involving them reaches your briefing.