UIST Submission

This is the audit to run in the final week before the Papers deadline. Numbers below are the verified 2026 cycle (checked 2026-07-08 via search renderings of uist.acm.org/2026/, direct fetch blocked): abstracts March 24, papers March 31, both 11:59 pm AoE. The 2026 deadline has passed — apply this skill to a 2027 submission only after re-reading that year's Author Guide, and expect the numbers to move.

What PCS collects, and when

UIST runs on PCS (Precision Conference Solutions): in the submission form, choose society "SIGCHI", conference "UIST 2026", track "UIST 2026 Papers". Two deadlines structure the week:

1. Abstract deadline (one week out). Title, abstract, author list, and metadata must exist in PCS. This gates reviewer assignment — a placeholder abstract that drifts from the final paper lands you with mismatched reviewers, the least fixable submission error.
2. Paper deadline. The PDF, the video figure, and all supplementary files. Everything reviewed arrives now; there is no later supplement slot.

Format gate

A short paper is a distinct category (smaller complete contribution), not a truncated standard paper — do not switch categories on deadline night to dodge the page count.

Anonymity, the UIST version

The standard is "reasonable effort," with two venue-specific twists:

• Third-person self-citation with full references. Do not redact your own citations; write about your prior work as if someone else built it.
• Overlap attachment. If a prior submission of yours overlaps significantly, upload an anonymized copy of it as supplementary material.

Sweep everything that ships: PDF text, PDF metadata, acknowledgements and grant numbers, repository and project-page URLs, video frames (lab posters, faces, institutional wallpaper), video audio (narrator self-identification), file names, and archive contents.

# Mechanical pass over the actual upload set
pdfinfo paper.pdf | grep -Ei 'author|creator|producer'
pdftotext paper.pdf - | grep -nEi 'github|gitlab|osf\.io|acknowledg|grant' | head
ffprobe -v error -show_entries format=duration:stream=width,height,codec_name video.mp4
unzip -l supplement.zip | grep -Ei '\.git|DS_Store|/Users/|/home/' | head

Deadline-week order of operations

1. T-7d: abstract deadline — final title/abstract in PCS, authors and conflicts complete, subject metadata chosen for reviewer routing.
2. T-5d: body frozen at the page limit; appendix absorbs the overflow.
3. T-4d: video locked (edits, captions, anonymization) and compressed to spec; verify PCS accepts the file size early, since caps were not published (待核实).
4. T-2d: full anonymity sweep on the merged PDF, video, and archive; run the mechanical checks above on the exact files to be uploaded.
5. T-1d: upload everything, then re-download from PCS and open cold — the re-downloaded copy is what reviewers see. Confirm the PCS abstract matches the PDF abstract word for word.
6. T-0: buffer only. AoE means UTC-12; convert once, write it on the wall.

PCS form fields worth drafting early

The form is read before the PDF and configured per track; draft its contents days ahead rather than composing in the upload session:

• Title and abstract — identical to the PDF, since the abstract drives reviewer matching from the abstract deadline onward.
• Author list and conflicts — complete at abstract time; adding authors later in the process is at minimum awkward and possibly restricted (待核实 per cycle).
• Subject/topic descriptors — choose for the reviewer pool you want; a fabrication paper tagged only "AI" gets AI reviewers.
• Prior-submission declarations — where the form asks about overlapping or previously reviewed material, answer literally; this is also where the anonymized overlap attachment gets referenced.
• Track category — standard vs short is a claim about contribution scale; reviewers hold you to the category you picked.

Last-day failure modes, ranked by frequency

Dual-track and dual-venue rules

• Papers under review elsewhere may not be under review at UIST at the same time.
• Within UIST 2026, the same work sent to both Demos and Posters is treated as a demo submission and rejected from Posters — pick one adjunct lane deliberately.
• A prior adjunct publication (demo/poster abstract) of the same system does not anonymize away: cite it in the third person and state the delta.

Output format

[Submission status] ready / blocked / at risk
[Format gate] pages <n>/<limit> · template ok? · video spec ok?
[Anonymity sweep] clean / leaks: <file: location>
[PCS state] abstract locked · uploads verified by re-download? 
[Remaining actions] <ordered, with owner and hour>