Senior Area Chair (SAC) Guidelines

Thank you for agreeing to serve as an SAC for NeurIPS 2021! This document contains an overview of your responsibilities and some guidelines for how to fulfill your role as an SAC.

As an SAC, your role is to oversee the work of a small number of ACs, making sure that the reviewing process goes smoothly. SACs serve as the first point of contact for ACs if they need assistance or guidance. SACs are responsible for helping ACs chase late reviewers, calibrating decisions across ACs, and discussing borderline papers. During the final decision-making phase, SACs discuss all proposed decisions with the program chairs (PCs).

Contact Info

If you encounter a situation that you are unable to resolve on your own, please contact the program chairs at neurips2021pcs@gmail.com. Any questions about conflicts of interest should go to the program chairs.

If the issue is related to OpenReview, email the OpenReview support team directly at neurips2021@openreview.net.

Main Tasks

  1. Preparation:

    • Please ensure that your preferred email address is accurate in your OpenReview profile. We will send most emails from OpenReview (i.e., noreply@openreview.net). Such emails are sometimes accidentally marked as spam. Please check your spam folder regularly. If you find such an email in there, please whitelist the OpenReview email address so that you will receive future emails from OpenReview.

    • Please log into OpenReview and make sure that your profile is up to date, so that you will be assigned relevant ACs to work with.

    • Read and agree to abide by the NeurIPS code of conduct.

    • Read what constitutes a conflict of interest for NeurIPS 2021.

    • In addition to the guidelines below, please familiarize yourself with the AC guidelines. You will be interacting significantly with ACs, so please make sure you understand what is expected of them.

  2. Bid on ACs: May

    • Log into OpenReview and bid on ACs you would like to work with. These bids will be used in the assignment process.

    • You will be assigned ~7 ACs to work with. When you receive your assignment, look it over carefully and email the PCs if you have any concerns.

  3. Help ACs with reviewer assignments if needed: Monday, June 14 – Monday, June 21

    • ACs will be reviewing and modifying reviewer assignments during this period. Your workload during this period should be light, but do make sure that each AC you work with has assigned four qualified reviewers for each paper (three are automatically assigned; ACs can replace them but must add a fourth reviewer manually). If an AC has difficulty finding a qualified reviewer not in conflict with the paper, please help them. Recall that ACs do not have access to author identities.

  4. Ensure that all papers have at least 3 quality reviews: Friday, July 16, 2021 – Tuesday August 3

    • Initial reviews are due Friday, July 16. Starting then, ACs should ensure that the reviewers have completed their reviews, send reminder emails if needed, and read all reviews to ensure they are up to standards. Again, your workload should be light during this period, but do check in to make sure that ACs are following up on missing reviews. Before the author response period starts, ensure that all of the papers your ACs are assigned have at least 3 high quality reviews. Make sure that ACs ensure that reviews are respectful to the authors. You are ultimately responsible for making sure the reviews are all there and high quality, so if an AC is unresponsive you will need to step in.

  5. Author response period: Tuesday, August 3 – Tuesday, August 10

    • During this period, reviews will be available to authors. If any reviews are still missing, it is urgent to help your ACs track them down or invite additional reviewers. SACs and ACs may invite external reviewers, but please do this through OpenReview directly rather than sharing submissions through other channels like email.

  6. Ensure that ACs initiate lead with reviewers: Tuesday, August 10 – Thursday, September 2

    • As soon as the author response is entered in the system, ACs should lead a discussion via OpenReview for each submission and make sure the reviewers engage in the discussion phase. If your assigned ACs have not initiated discussions, prompt them to do so.

    • Remind ACs to submit preliminary meta-reviews for each paper by September 2 and begin to schedule one-on-one discussions with them for the next phase. Consider organizing meetings between pairs or groups of ACs so that they have the chance to talk over their decisions and better calibrate; you may choose to be present for these meetings or ask that ACs meet with each other before discussing papers with you. Check carefully for COIs before asking ACs to discuss papers with each other.

  7. SACs discuss papers with ACs and make initial accept/reject decisions: Thursday, September 2 – Thursday, September 9

    • Help calibrate decisions by working closely with your ACs. Schedule meetings with them individually and/or in groups. We will provide you with templates you can use to organize the discussion. Pay particularly close attention to borderline papers and papers in which the AC’s recommendation goes against the recommendations of the reviewers.

    • If you feel that a particular AC needs your guidance, please read all reviews for papers assigned to them. Make sure they are respectful and acknowledge the authors’ response.

    • Read all meta-reviews. Make sure they explain paper decisions to the authors. Meta-reviews should augment the reviews, and explain how the reviews, author responses, and discussion were used to arrive at the decision. Dismissing or ignoring a review is not acceptable unless there is a good reason for doing so.

  8. Meet with program chairs to finalize decisions: Friday, September 10 – Friday, September 24

    • Be prepared to discuss all borderline papers and cases in which the recommendation of the AC goes against the recommendations of the reviewers.

    • Update meta-reviews to accurately reflect the final decision.

  9. Notification: Tuesday, September 28

Best Practices

  • Be responsive. Respect deadlines and respond to emails as promptly as possible. Make sure that your preferred email address is accurate in your OpenReview profile and that emails from noreply@openreview.net don’t go to spam. If you will be unavailable (e.g., on vacation) for more than a few days—especially during important windows (e.g., decision-making)—please let the program chairs know as soon as possible.

  • Be proactive. It is your responsibility to ensure that the review process goes smoothly. Check in to make sure that the ACs you work with are responsive, help them find emergency reviewers, and make sure discussion is happening on their papers.

  • Be kind. It is important to acknowledge that personal situations, in particular during this year of a global pandemic, may lead to late or unfinished work among reviewers and ACs. In the event that a reviewer or an AC is unable to complete their work on time, we encourage you to be considerate of the personal circumstances; you might have to pick up the slack in some cases. If necessary, make a back-up plan with another reviewer or AC, and be flexible to the extent possible. In all communications, exhibit empathy and understanding.

  • Respect conflicts of interest. Since the reviewing process is double blind at the level of ACs, it is your responsibility to be on the lookout for uncaught conflicts of interest. If you notice a conflict of interest with a submission that is assigned to one of your ACs, contact program chairs right away. Do not talk to other SACs about submissions assigned to your ACs without prior approval from program chairs since other SACs may have conflicts with these submissions. Do not talk to other SACs or ACs about submissions you are an author on or submissions with which you have a conflict of interest.

  • Familiarize yourself with the code of conduct. All participants must agree to abide by the NeurIPS code of conduct.

Confidentiality

You must keep everything relating to the review process confidential. Do not use ideas, code, or results from submissions in your own work until they become publicly available (e.g., via a technical report or a published paper for ideas/results, via open source for code). Do not talk about or distribute submissions (whether it is the code, or the ideas and results described in them) to anyone without prior approval from the program chairs. Code submitted for reviewing cannot be distributed. If you wish to invite an external reviewer, do so through OpenReview rather than sharing submissions through another channel.

Quick Links