\documentclass{article}

% --- Venue style (swap per target) ---
\usepackage[preprint]{neurips_2024}

% --- Core packages ---
\usepackage[utf8]{inputenc}
\usepackage[T1]{fontenc}
\usepackage{hyperref}
\usepackage{url}
\usepackage{booktabs}       % Professional tables
\usepackage{amsfonts}
\usepackage{amsmath}
\usepackage{amssymb}
\usepackage{mathtools}       % dcases, coloneqq
\usepackage{nicefrac}
\usepackage{microtype}       % Better typography
\usepackage{xcolor}
\usepackage{graphicx}
\usepackage{subcaption}      % Subfigures
\usepackage{algorithm}
\usepackage{algorithmic}
\usepackage{pgfplots}
\usepackage{tikz}
\pgfplotsset{compat=1.18}

% --- Math operators ---
\DeclareMathOperator*{\argmin}{arg\,min}
\DeclareMathOperator*{\argmax}{arg\,max}
\DeclareMathOperator{\E}{\mathbb{E}}
\newcommand{\R}{\mathbb{R}}
\newcommand{\loss}{\mathcal{L}}
\newcommand{\dataset}{\mathcal{D}}
\newcommand{\param}{\boldsymbol{\theta}}
\newcommand{\norm}[1]{\left\lVert #1 \right\rVert}

\title{Your Title Here}
\author{Anonymous}  % Submission phase

\begin{document}
\maketitle
\begin{abstract}
Your abstract (NeurIPS: max 250 words).
\end{abstract}

\begin{algorithm}[t]
\caption{Stochastic Gradient Descent with Momentum}
\label{alg:sgd-momentum}
\begin{algorithmic}[1]
\REQUIRE Learning rate $\eta$, momentum $\beta$, initial params $\param_0$
\ENSURE Trained parameters $\param_T$
\STATE $\mathbf{v}_0 \leftarrow \mathbf{0}$
\FOR{ = 1$ \TO $}
    \STATE Sample mini-batch  \sim \dataset$
    \STATE $\mathbf{g}_t \leftarrow \nabla_{\param} \loss(\param_{t-1}; B_t)$
    \STATE $\mathbf{v}_t \leftarrow \beta \mathbf{v}_{t-1} + \mathbf{g}_t$
    \STATE $\param_t \leftarrow \param_{t-1} - \eta \mathbf{v}_t$
\ENDFOR
\RETURN $\param_T$
\end{algorithmic}
\end{algorithm}

\begin{figure}[t]
  \centering
  \begin{subfigure}[b]{0.48\textwidth}
    \centering
    \includegraphics[width=\textwidth]{figures/loss_curve.pdf}
    \caption{Training loss}
    \label{fig:loss}
  \end{subfigure}
  \hfill
  \begin{subfigure}[b]{0.48\textwidth}
    \centering
    \includegraphics[width=\textwidth]{figures/accuracy.pdf}
    \caption{Validation accuracy}
    \label{fig:acc}
  \end{subfigure}
  \caption{Training dynamics on CIFAR-10. (a) Loss converges within 50 epochs.
           (b) Accuracy plateaus at 94.2\%.}
  \label{fig:training}
\end{figure}

\begin{figure}[t]
\centering
\begin{tikzpicture}
\begin{axis}[
    width=0.75\columnwidth,
    height=5cm,
    xlabel={Epoch},
    ylabel={Test Accuracy (\%)},
    legend pos=south east,
    grid=major,
    legend style={font=\small},
]
\addplot[blue, thick, mark=none] table[x=epoch, y=ours, col sep=comma]
    {data/results.csv};
\addplot[red, dashed, thick, mark=none] table[x=epoch, y=baseline, col sep=comma]
    {data/results.csv};
\legend{Ours, Baseline}
\end{axis}
\end{tikzpicture}
\caption{Comparison against baseline on ImageNet.}
\label{fig:comparison}
\end{figure}

\begin{table}[t]
\caption{Results on standard benchmarks. Best in \textbf{bold}, second \underline{underlined}.}
\label{tab:results}
\centering
\small
\begin{tabular}{lcccc}
\toprule
Method & CIFAR-10 & CIFAR-100 & ImageNet & Params (M) \\
\midrule
ResNet-50        & 95.1 & 78.3 & 76.1 & 25.6 \\
ViT-B/16         & 96.2 & 82.1 & 77.9 & 86.4 \\
Baseline         & 95.8 & 80.5 & 77.2 & 24.1 \\
\midrule
\textbf{Ours}    & \textbf{97.1} & \textbf{84.3} & \underline{78.8} & 26.3 \\
Ours (small)     & \underline{96.5} & \underline{83.1} & \textbf{79.2} & 12.8 \\
\bottomrule
\end{tabular}
\end{table}

% In preamble (natbib usually loaded by venue style)
\usepackage[numbers,sort&compress]{natbib}
\bibliographystyle{plainnat}

% At end of document
\bibliography{references}

@inproceedings{vaswani2017attention,
  title     = {Attention is All You Need},
  author    = {Vaswani, Ashish and Shazeer, Noam and Parmar, Niki and
               Uszkoreit, Jakob and Jones, Llion and Gomez, Aidan N and
               Kaiser, {\L}ukasz and Polosukhin, Illia},
  booktitle = {Advances in Neural Information Processing Systems (NeurIPS)},
  volume    = {30},
  year      = {2017}
}

@article{devlin2019bert,
  title   = {{BERT}: Pre-training of Deep Bidirectional Transformers
             for Language Understanding},
  author  = {Devlin, Jacob and Chang, Ming-Wei and Lee, Kenton and
             Toutanova, Kristina},
  journal = {arXiv preprint arXiv:1810.04805},
  year    = {2019}
}

% After \bibliography{references} in main document
\newpage
\appendix
\section{Proof of Theorem 1}
\label{app:proof}
...

\section{Additional Experiments}
\label{app:experiments}

\section{Hyperparameter Details}
\label{app:hyperparams}

\section{Compute Resources}
\label{app:compute}