Loading
A result is reproducible when someone else can get it again from the same data and code; a negative result is an experiment that did not show the hoped-for effect, and reporting it is how a field stops repeating dead ends.
Reproducibility, closely related to replicability, is a major principle of the scientific method: results should be achieved again with high reliability (Wikipedia). A null result is one without the expected content, which does not support the hypothesis (Wikipedia). In cancer AI the published record is skewed towards wins, so frozen-foundation-model results that fail to beat a baseline, panels that wash out a gain or pretraining at scale that adds nothing are worth recording as findings, with the seeds, versions and checksums that let them be rerun.
Shares CITATION.cff (Citation File Format), Zenodo DOIs for data and code, Cancer AI vocabulary (CanSim terms map) and the tag cansim-terms.
Shares Model cards and datasheets for datasets, Zenodo DOIs for data and code, Cancer AI vocabulary (CanSim terms map) and the tag cansim-terms.
Shares Pre-registered experiment, Cancer AI vocabulary (CanSim terms map) and the tag cansim-terms.
Shares Model cards and datasheets for datasets, Cancer AI vocabulary (CanSim terms map) and the tag cansim-terms.
Shares Benchmarks, leaderboards and contamination, Cancer AI vocabulary (CanSim terms map) and the tag cansim-terms.
Shares Model cards and datasheets for datasets, Cancer AI vocabulary (CanSim terms map) and the tag cansim-terms.
Shares Provenance fields for research data, Cancer AI vocabulary (CanSim terms map) and the tag cansim-terms.
Shares Provenance fields for research data, Cancer AI vocabulary (CanSim terms map) and the tag cansim-terms.