AI 服務暫時不可用，以下為來源正文，待恢復後補全翻譯。
What does it mean for data to be AI ready? By A. Sina Booeshaghi · August 29, 2026 That is the central question of AIDaR: AI Data Readiness for Scientific Discovery, a workshop I am co-organizing at NeurIPS in Paris this year with Michal Rosen-Zvi, Zoe Piran, Kenny Workman, Edaeni Hamid, Vladimir Ermakov, and Arindam Sett. We are accepting submissions until September 5, 2026 at 11:59 p.m. AoE. Workshop submissions follow the standard NeurIPS process. Authors submit a manuscript through OpenReview and reviewers respond there. Much of the code behind scientific papers is already managed through GitHub. Authors at some conferences use Anonymous GitHub to share a read-only, anonymized copy of their repository. These repositories are useful for blinded peer review, but reviewers cannot take advantage of GitHub tools such as issues or pull requests to give targeted feedback on the work. This limitation motivated a new submission system called submityour.work. This system tests whether GitHub's review tools lead to more useful and reproducible feedback for workshop submissions. It also tests whether a consistent, machine-readable project is easier for people and agents to use. submityour.work allows authors to connect their OpenReview submission to a private GitHub repository for review: Submission: After submitting to OpenReview, authors submit their project folder using an AI agent, the command-line tool, or a browser. Authors do not need a GitHub account. Redaction: The AIDaR submission tool removes Git history, checks a temporary copy for names and common secret patterns, and sends the files to a private GitHub repository. The server does not retain the uploaded copy. Review: Reviewers inspect the files and leave feedback through GitHub issues and pull requests. Authors can use a stable submission ID to submit revisions and respond through the same system. submityour.work is a prototype we built for this workshop. Participation is entirely opt-in and does not affect your chances of acceptance. We developed it to learn what an AI-data-ready workshop submission might require. If authors use it, we can learn what works, what fails, and which safeguards are missing. We expect to make improvements. If you have questions or feedback, please feel free to email [email protected]. What happens after decisions are made? Once decisions are made, authors who submitted their work through the pilot can ask us for a copy of their repository or request a transfer to their GitHub account. A transferred repository will include reviewer comments or contributions only with the reviewers' consent. We will delete all remaining repositories after the workshop.