Run this audit before uploading a learning-theory paper to COLT. Every number below was checked against the COLT 2026 Call for Papers on 2026-07-08; reopen the current CFP at learningtheory.org before trusting any of it in a later cycle, because COLT republishes its rules from scratch each year.
https://cmt3.research.microsoft.com/COLT2026. COLT has used other systems in the
past, so never assume the portal — follow the link in the live CFP.anon option suppresses author names.Because the appendix rides inside the same PDF as the body, COLT has no separate supplement portal to forget — but also no second chance to patch a broken proof file. Treat the appendix as reviewable text, not as an attachment:
The exact style file name changes with the year (e.g., a colt2026 package for the
2026 cycle). Shape, not filename, is the stable part:
\documentclass[12pt]{article}
\usepackage[anon]{colt2026} % 'anon' hides authors for review; drop it at camera-ready
\usepackage{times}
\title[Short Title]{Tight Rates for a Learning Problem}
\coltauthor{\Name{Anonymous Author} \Email{anon@example.org}}
\begin{document}
\maketitle
\begin{abstract}%
One-paragraph statement of the problem, the main theorem, and the technique.
\end{abstract}
% <= 12 PMLR-formatted pages of body
% \bibliography{refs} -- references excluded from the limit
% \appendix -- full proofs; no page limit, same PDF
\end{document}
Download the real template from the current CFP; do not reuse a prior year's file, and never alter margins, font size, or spacing — layout tampering is a desk-reject risk at any PMLR venue.
The COLT 2026 CFP barred submissions substantially similar to papers previously published, accepted for publication, or in parallel submission to other peer-reviewed conferences with proceedings or to journals. It also barred sending a paper under COLT review to a journal until decisions are released or the paper is withdrawn.
| Situation | COLT 2026 reading | Safe action |
|---|---|---|
| Same results on arXiv | Preprints are standard in the theory community | Post freely; do not cite your arXiv version in a way that de-anonymizes |
| Under review at a journal | Forbidden overlap | Withdraw one venue before the COLT deadline |
| Extended abstract at a proceedings venue (e.g., STOC/FOCS/ALT) | Forbidden if substantially similar | Differentiate results substantially or pick one venue |
| Non-archival workshop talk | Generally tolerated in theory culture | Verify the workshop is truly non-archival, then declare if the form asks |
| Journal submission planned after COLT decision | Explicitly allowed once decisions are out | Calendar the journal submission after notification |
| Trigger | Severity | Window to repair |
|---|---|---|
| Body beyond 12 PMLR pages | Desk reject | Only before the deadline |
| Author names or acknowledgements left in | Desk reject or chair flag | Only before the deadline |
| Parallel-submission violation | Rejection at any stage, discoverable late | Withdraw the other venue now |
| Main theorem provable only via a gap-ridden appendix | Review-stage rejection | Fix during the proof-verification pass |
| Missing formal problem setup in the body | Review-stage damage | Fixable in the writing pass |
\thanks, metadata, repository links, and
self-citation phrasing ("in our earlier work [3]" must become third person).[COLT readiness] Ready / Needs fixes / Not ready
[Blocking checks] <CMT record / 12-page body / single PDF / anonymity / parallel submission>
[Proof-integrity risk] <statement-proof drift, unverified lemma, external-result misuse>
[Desk-reject risk] <one issue>
[Fix order] <ordered actions before the AoE deadline>