MAINT: Attribute in-the-wild jailbreak templates to paper author affiliations#2276
Open
romanlutz wants to merge 6 commits into
Open
MAINT: Attribute in-the-wild jailbreak templates to paper author affiliations#2276romanlutz wants to merge 6 commits into
romanlutz wants to merge 6 commits into
Conversation
Align dataset authors, affiliations, citation details, and YAML provenance with authoritative paper and source metadata. Co-authored-by: Copilot App <223556219+Copilot@users.noreply.github.com> Copilot-Session: dd1ad5b0-f48b-4a55-b75e-0b875d438047
Co-authored-by: Copilot App <223556219+Copilot@users.noreply.github.com> Copilot-Session: dd1ad5b0-f48b-4a55-b75e-0b875d438047
Add the CISPA Helmholtz Center for Information Security and NetApp affiliations to the 488 TrustAIRLab in-the-wild jailbreak templates, upgrade the `Do Anything Now` citation from arXiv preprint to the CCS 2024 proceedings, and update the provenance test to assert the full set of curator/paper groups. Co-authored-by: Copilot App <223556219+Copilot@users.noreply.github.com> Copilot-Session: b04dd4fc-cf78-464d-91d8-351ca08d90c5
Drop the `TrustAIRLab` curator label from the `groups` list of the 488 in-the-wild jailbreak templates, leaving just the paper author affiliations (CISPA Helmholtz Center for Information Security and NetApp). TrustAIRLab remains recorded as the curator in each template's description, source URL, and metadata, and the provenance test still asserts it. Co-authored-by: Copilot App <223556219+Copilot@users.noreply.github.com> Copilot-Session: b04dd4fc-cf78-464d-91d8-351ca08d90c5
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
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
Description
Corrects the
groupsattribution on the 488 TrustAIRLab In-The-Wild Jailbreak Prompts templates so they credit the paper's author affiliations instead of the curating-lab label:groupschanged from[TrustAIRLab]to[CISPA Helmholtz Center for Information Security, NetApp]— the affiliations of the "Do Anything Now" authors (Shen et al., CCS 2024).TrustAIRLab, the curating lab (housed within CISPA), remains recorded as the curator in each template'sdescription,sourceURL, andmetadata.source_dataset, so no provenance is lost.shen2023donotanythingentry indoc/references.bibfrom the arXiv preprint to the published CCS 2024 proceedings (adds DOI, pages, and the official dataset note).Tests and Documentation
tests/unit/datasets/test_jailbreak_in_the_wild.pyto assert the two-institutiongroupslist; the test still assertsTrustAIRLabappears in each template's description, so the curator stays documented.pytest tests/unit/datasets/test_jailbreak_in_the_wild.py→ 3427 passed.