-
Update README.md
-
Drop the superseded ESM-C metrics build
-
Card: strict pre-cutoff corpus, de-anonymised repo name
-
ESM-C scores over the strict pre-cutoff feature corpus
-
Punctuation fix
-
Drop the superseded comparability note
-
List the Zhang configs in the loading section
-
Dataset card revisions
-
Remove the superseded ESM-C build
-
Document the pre-cutoff ESM-C corpus and the Zhang et al. set
-
Add ESM-C scores over the Zhang et al. set
-
Add the Zhang et al. set sequences
-
Add the Zhang et al. scored pair set
-
ESM-C scores over both feature corpora
-
Card wording pass on the two metrics sections
-
Publish ESM-C scores for both benchmark sets
Same 24,022 pairs as esmfold2_metrics. Pairs with no active features are scored 0 and keep their missing_features_a/b status.
-
Trim the join note and the size-budget note from the metrics section
-
Publish ESMFold2 scores for both benchmark sets in one file
Replaces esmfold2_leakage_free_metrics with esmfold2_metrics: the leakage-free and leaked arms concatenated under a set column. -
Anonymize dataset card for review: drop Paper TODO, add code link
-
Add Flock v1 dataset files and dataset card
-
Update README.md
-
initial commit
