Where one study disagrees with what other studies report for the same models, and what we are doing to resolve each case.
These are automatic leads, not findings of error. Each release, a check compares how every study ranks the models it shares with other studies. It flags a study that reverses the order the others agree on, or whose gap to a shared reference model lies well outside theirs. Different tasks, populations or metrics often explain a flag.
A flag never hides a result. Flagged results stay in every table, page and download, marked Contested. A result is withheld only when a reviewer marks it disputed.
ProteinGym: Large-Scale Benchmarks for Protein Fitness Prediction and Design ranks VESPA above CARP in 18 of 19 comparisons; Livesey and Marsh 2025, Genome Biology 26:104 (Europe PMC full text XML) ranks CARP above VESPA in 6 of 6 comparisons. No other study gives a verdict, so neither is the outlier.
2 papers rank CARP above VESPA in 7 of their 25 comparisons and VESPA higher in 18. Of those papers, ProteinGym ranks VESPA higher on its own (18 of 19).
ProteinGym: Large-Scale Benchmarks for Protein Fitness Prediction and Design ranks VESPA above DeepSequence in 13 of 17 comparisons; Livesey and Marsh 2025, Genome Biology 26:104 (Europe PMC full text XML) ranks DeepSequence above VESPA in 5 of 6 comparisons. No other study gives a verdict, so neither is the outlier.
2 papers rank DeepSequence above VESPA in 9 of their 26 comparisons and VESPA higher in 14; 3 are ties. Of those papers, ProteinGym ranks VESPA higher on its own (13 of 19).