Authors Beware: Common Submission Problems

We reserve the right to desk reject the submissions that violate the requirements stated in the ARR call for papers. This checklist supplements the call for papers, and it is intended to help the authors check for the common problems prior to submission. Too many authors each cycle find themselves frustrated with desk rejections over small things that they could easily fix. We wrote this checklist to help you avoid that situation – especially if this is your first submission.

Please check all these points in advance, ideally at least a week. Some of the chese points require coordination with other authors or other people in your research group, and/or may remind you that some substantial changes are necessary in the paper.

Scope of submissions

  • The paper follows the scope of ARR CFP. ARR is a reviewing venue that mainly serves the conferences of the Association for Computational Linguistics, and the submissions are expected to make a contribution to the computational processing of language.

Author and reviewer registration

  • All authors are included, and in the final order. No changes to author list or its order will be allowed after submission. Please consult the guidelines on authorship criteria.
  • All authors completed the OpenReview profiles prior to submission. You may refer to this tutorial.
  • The author who is qualified and nominated to review has (prior to submission):
    • filled in Semantic Scholar, DBLP, and ACL Anthology profiles, and updated their listed papers to reflect their current research interests.
    • indicated their availability in this reviewing cycle, ideally reciprocating the amount of effort that their submission(s) will receive (each submission receives at least 3 reviews and a metareview, plus effort from chairs and any emergency reassignments).
    • if you are a member of a large research group that submits multiple papers in the same ARR cycle: make sure that you are nominating different senior co-authors. If there are too few qualified reviewers in the group, and their nomination on multiple papers is unavoidable, they should increase their review load, and undertake to mentor secondary reviewers. Again, each submitted paper receives at least 4 reviews incl. meta-review, and the system is not sustainable if some groups receive a lot more volunteer effort than they contribute.

ARR recognizes that there may be situations that qualify for an exemption from the reviewing load (e.g. if no authors are qualified to review, or if the qualified author is already serving at ARR in some other capacity). The form also allows to specify your justification for requesting an exemption.

Dual submissions and resubmissions

  • The paper is not and will not be under review at any archival venue while it is reviewed at ARR.
  • Only one copy of this paper is submitted. Please check with the other authors that nobody accidentally submitted twice.
  • If the paper is presented as a new submission, it must be on a sufficiently different topic that any prior reviews no longer apply. Simply reframing the paper, further improving the methods, and adding more baselines or analysis does not count.
  • If the paper is a revision, the submission follows the resubmissions policy in the CFP: the previous version must be acknowledged, and you should provide a summary of changes and revisions. You can also provide justifications for not performing some of the requested changes or even requesting new AC/reviewers.
  • Optionally, you can include a color-coded version of the paper that highlights the revisions, in the same pdf as the revision notes. It may help the AC and reviewers to focus on assessing the changes.

Paper Format

  • The paper follows page limits (8 for long, 4 for short, not including limitations and ethics sections).
  • The paper has a section titled “Limitations” after the conclusion (not in the appendices).
  • The Limitations section only includes the discussion of limitations, and no new experiments, figures or analysis.
  • The paper follows ACL formatting. You are encouraged to use the ACL pubcheck tool to check for common problems. There is an official Overleaf template. Note that this template has a [review] setting that must be on for the submission.
  • The limitations section, the optional ethical considerations section, references, and appendices should be included in the pdf for the paper (not counting towards the page limit), and not be submitted as a separate PDF.
  • There are no comments or any leftover meta-text from prior revisions.

Note that any supplemental materials (code, data, appendices) should be truly supplemental; a reader should be able to understand the work from the paper alone.

Anonymization and Citation

  • The submission must not contain explicit references to the authors’ prior work or otherwise disclose their identity.
  • Any supplementary materials are also properly anonymized.
  • The acknowledgements section is not included.
  • There are no links to non-anonymous repositories or services like Dropbox that may track who opened the link.
  • The anonymity status set at submission time is binding. If you are not sure whether you will preprint the paper during the review period, please select “we are considering…” option.

Please refer to this page for the ACL policies for review and citation. Most of this information is also covered in the template and format guidelines linked above.

Responsible NLP Research

Responsible NLP research means (among other things) taking into account issues related to ethics and to reproducibility. Our submission form includes the responsible NLP research checklist (detailed explanation of all its questions is available here). Note, prior to February 2024, this was submitted as a separate PDF. Besides reminding the authors about the best practices, it also serves as a kind of FAQ for ACs and reviewers: they can get the information they need without spending the author response time and effort. At ACL’25 the checklist will be published as an appendix to accepted papers, and starting from December 2024 inappropriately filled checklist can be considered grounds for desk rejection. For example, some authors provide misleading responses, or just check “yes” even for the questions that are not relevant for their paper.

Authors are encouraged to include data sheets and model cards in supplemental material, as appropriate. They are also encouraged to devote a section of their paper to concerns of the ethical impact of the work and to a discussion of broader impacts of the work, which will be taken into account in the review process. This discussion may extend into a 5th page (short papers) or 9th page (long papers).