Artificial Intelligence and Statistics 2023

1. I missed the abstract submission deadline. Can I still submit a full paper? No. You must submit the abstract in time for the deadline on October 6th.

2. While submitting a paper, do we need to nominate one of the authors as a reviewer? While not mandatory, we encourage authors to also serve as reviewers in order to ensure that all submissions can receive a sufficient number of reviews whilst ensuring an acceptable load for each reviewer. Due to the recent increase in submissions, we expect we will need a significantly larger number of reviewers than in previous years. Your nominations are important in order to maintain the quality of the conference. There will be a section in the paper submission form to nominate any of the authors to be reviewers. Nominated authors will receive an invitation to be a reviewer, which can be accepted or declined.

3. We want to submit a version of the work to arXiv. Is it allowed? Yes, submission to arXiv is allowed. However, keep anonymity in the AISTATS submission. Do not cite the arXiv paper, for example.

4. Can we advertise the preprint of my submission on social media? We strongly discourage advertising the preprint on social media or in the press while under submission to AISTATS. Under no circumstances should your work be explicitly identified as AISTATS submission at any time during the review period, i.e., from the time you submit the abstract to the communication of the accept/reject decisions.

5. Can the author list be changed? You can change the author list until the deadline of full paper submission (October 13th). After this point, only changes to the author order are allowed, but new authors won’t be added.

6. Can I withdraw my submission? Yes. You can withdraw your submission at any point in the reviewing process. For statistical purposes, we will count a submission as “rejected” if it is withdrawn after reviews have been made available to authors.

7. Can we revise the abstract after the abstract deadline? You can revise your abstract before the full paper deadline, but the changes must be minor (e.g., fixing typos or clarify/correct wording). If the contents or length of the abstract is significantly altered, the submission may be desk-rejected.

8. I am an author as well as a reviewer/area chair, but I can only find my author (or reviewer) role in CMT. What should I do? On the CMT page, you will find a “Select Your Role” tab on the top right. If you click on the current role (e.g., Author), you will see a drop down menu showing your other roles, and you can pick the one you need.

9. Can I use a different email for my CMT reviewer account and TPMS account? The easiest approach is to have the same email for your CMT account and the TPMS account. Alternatively, you may use the new “Linked Account” feature of CMT (see “Link to Account” in the dropdown list at the top right of the console), to link your Area Chair/Reviewer CMT account with another CMT account that uses the same email as your TPMS account. Regardless of the chosen path, it will be critical to have CMT and TPMS accounts with matched email addresses.

10. If our paper is a resubmission from a previous conference, do I have to indicate so in the submission form? Yes, authors must specify whether the paper is a resubmission and, if so, they must summarize the main criticisms raised by the previous reviewers and how these have been addressed. This information will be visible to Area Chairs only, but not to reviewers.

11. Can we submit code/dataset with our submission? Yes, and authors are encouraged to do so, as long as anonymity is preserved. For example, you may include your code as part of the supplementary material or through an anonymized link. Please note that papers that promise to release code or dataset (either at submission time or during the rebuttal phase) will be automatically rejected if the authors fail to make the code/dataset public by the camera-ready deadline.

12. Does the paper need to discuss the method assumptions and limitations? It is not compulsory, but authors are nonetheless encouraged to discuss these points, as they will be positively considered during the review phase.

13. Does the paper need to discuss the societal impact? It is not mandatory, but authors are encouraged to discuss the societal implications of their research.

14. Is there a reproducibility checklist that my paper needs to comply with? Authors are encouraged to follow the suggested checklist provided at the end of this document.

15. Will the questions about industry/academic paper or hardware usage be visible to reviewers or affect the paper decision? No, this information will not be visible to reviewers or ACs and will not be taken into account in the paper decisions. It will only be used for statistical purposes.

16. How can we obtain the citations (references) of our paper in the (Author, Year) format? One option to achieve that goal is to use the latex package natbib. For that, you may uncomment lines 19-21 and 24 at the beginning of your main TEX file (sample_paper.tex in the AISTATS2023 Paper Pack). In other words, you can use the following:

% If you use natbib package, activate the following three lines:
\usepackage[round]{natbib}
\renewcommand{\bibname}{References}
\renewcommand{\bibsection}{\subsubsection*{\bibname}}

% If you use BibTeX in apalike style, activate the following line:
\bibliographystyle{apalike}

\bibliography{your_bib_file}


Note that, when using natbib, you can add or remove the parentheses as follows:

\cite{foo2021}   % produces Foo et al. (2021)
\citep{foo2021}  % produces (Foo et al., 2021)


## AISTATS 2023: Guidelines and Checklist

Below is a suggested guideline and checklist for paper submissions:

1. For all models and algorithms presented, check if you include:
1. A clear description of the mathematical setting, assumptions, algorithm, and/or model.
2. An analysis of the properties and complexity (time, space, sample size) of any algorithm.
3. (Optional) Anonymized source code, with specification of all dependencies, including external libraries.
2. For any theoretical claim, check if you include:
1. A statement of the result.
2. A clear explanation of any assumptions.
3. A complete proof of the claim.
3. For all figures and tables that present empirical results, check if you include:
1. A complete description of the data collection process, including sample size.
3. An explanation of any data that were excluded, description of any pre-processing step.
4. An explanation of how samples were allocated for training / validation / testing.
5. The range of hyper-parameters considered, method to select the best hyper-parameter configuration, and specification of all hyper-parameters used to generate results.
6. The exact number of evaluation runs.
7. A description of how experiments were run.
8. A clear definition of the specific measure or statistics used to report results.
9. Clearly defined error bars.
10. A description of results with central tendency (e.g., mean) & variation (e.g., stddev).
11. A description of the computing infrastructure used.