neurips_2026.sty — Official NeurIPS 2026 Style File

What is neurips_2026.sty?

neurips_2026.sty is the official LaTeX style file for the 2026 Neural Information Processing Systems conference. It controls every aspect of how a NeurIPS paper is typeset: page geometry, fonts, section headings, the bibliography style, and the all-important \ack environment that hides acknowledgments during anonymous review.

You can't submit to NeurIPS without it. The conference review system rejects papers that don't conform to the official style, and the easiest way to conform is to drop neurips_2026.sty into your project and \usepackage{neurips_2026}.

Quick usage

\documentclass{article}

% For anonymous submission (default):
\usepackage{neurips_2026}

% For camera-ready (final) version:
% \usepackage[final]{neurips_2026}

\title{Your Paper Title}
\author{%
  Author One \\
  Affiliation \\
  \texttt{author@domain} \And
  Author Two \\
  Affiliation \\
  \texttt{author2@domain}
}

\begin{document}
\maketitle
\begin{abstract}
  Your abstract.
\end{abstract}

% ... your paper ...

\begin{ack}
  This work was supported by ...
\end{ack}

\bibliographystyle{plainnat}
\bibliography{references}
\end{document}

Submission mode vs final mode

The single most important thing neurips_2026.sty does is switch between two modes:

• Submission mode (default, \usepackage{neurips_2026}) hides author names, affiliations, and the entire \ack block so the paper is anonymized for double-blind review.
• Final mode (\usepackage[final]{neurips_2026}) shows authors and acknowledgments — for the camera-ready version after your paper is accepted.

If reviewers see your name in submission mode, your paper can be desk-rejected for de-anonymization. The style file handles this for you — don't hardcode author visibility.

What about the bibliography style?

neurips_2026.sty sets up a bundled .bst file that produces the NeurIPS-specific numbered citation format. You can use it directly with \bibliography{refs} or override with \bibliographystyle{plainnat} if you prefer natbib.

Frequently asked questions