fix: scope text-structure fingerprint (id 43) robustness claim to match benchmark - #60
Merged
domguinard merged 1 commit intoJul 22, 2026
Conversation
domguinard
approved these changes
Jul 21, 2026
Collaborator
|
Approved during the WM TF meeting of July 22nd. |
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
2 participants
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Corrects the robustness description of
com.writerslogic.text-structure.1(id 43) to match measured behavior.The prior wording ("survives synonym-level paraphrase"; "robust to synonym substitution") overstates paraphrase robustness. The fingerprint hashes sentence-length sequences, so it is invariant only to substitutions that preserve token and sentence lengths. A reproducible benchmark on the PAN'26 text corpus (Zenodo 18620130) confirms this: balanced accuracy 0.99 under light length-preserving synonym substitution, dropping to chance (0.52) under substantive lexical paraphrase (DIPPER) that changes token counts.
The revised text scopes the claim to light, length-preserving synonym substitution and adds substantive lexical paraphrase to the non-robust list. Description text only — no schema or functional change.
Methodology and reproducible harness: https://github.com/writerslogic/c2pa-text-binding/blob/main/ROBUSTNESS.md (benchmark: https://github.com/writerslogic/c2pa-text-binding/blob/main/examples/robustness_bench.rs).