A static, single-page conference schedule generated from the accepted-paper assignment workbook and submission-status export.
The checked-in data/papers.json is the deployable data file. To rebuild it, run:
node scripts/build-data.mjsThe script cross-references accepted submission numbers and forum links, calculates each 15-minute oral slot from the confirmed presentation order, and enriches exact public arXiv title matches with author names.
Withdrawn submission numbers are listed in data/withdrawn-papers.json and excluded during rebuilds. The deployable schedule currently contains 91 papers after nine removals.
Organizer access can resolve the author list and profile link directly from every submission note. Install the official client, then run the importer; it prompts for credentials securely and does not save them:
python3 -m pip install openreview-py
python3 scripts/import-openreview-authors.py --dry-run
python3 scripts/import-openreview-authors.py
node scripts/verify-data.mjsRun this after build-data.mjs, because rebuilding the paper data replaces data/papers.json.
Equal-contribution designations are kept in data/author-notes.json and copied into data/papers.json during a rebuild. The author importers preserve these paper-level notes.
If the camera-ready audit workbook is available, it can populate every author name without API credentials while preserving its first-author OpenReview profile links:
node scripts/import-authors-workbook.mjs "/path/to/MLSP_2026_First author student status.xlsx"
node scripts/verify-data.mjsServe the directory over HTTP (the page fetches its JSON data file):
python3 -m http.server 4173Then open http://localhost:4173/.