MetivtaEval is built around a community-governed standard for AI-driven Torah scholarship. We welcome contributions to our public dataset and rubrics.
All changes are managed through GitHub Issues and Pull Requests.
- Proposal: To suggest a new question, a better ground-truth answer, or a change to a rubric, please open a new GitHub Issue.
- Community Polling: We use emoji reactions to gauge community support. If you agree with a proposal, please add a 👍 reaction to the original issue post.
- Review Threshold: Any proposal that receives 50 or more 👍 reactions within 21 days will be formally reviewed by the project maintainers.
- Decision: After review, a maintainer will either create a Pull Request to implement the change or post a comment explaining why it was not accepted.