@Article{Faster2023,
author={Mankowitz, Daniel J.
and Michi, Andrea
and Zhernov, Anton
and Gelmi, Marco
and Selvi, Marco
and Paduraru, Cosmin
and Leurent, Edouard
and Iqbal, Shariq
and Lespiau, Jean-Baptiste
and Ahern, Alex
and K{\"o}ppe, Thomas
and Millikin, Kevin
and Gaffney, Stephen
and Elster, Sophie
and Broshear, Jackson
and Gamble, Chris
and Milan, Kieran
and Tung, Robert
and Hwang, Minjae
and Cemgil, Taylan
and Barekatain, Mohammadamin
and Li, Yujia
and Mandhane, Amol
and Hubert, Thomas
and Schrittwieser, Julian
and Hassabis, Demis
and Kohli, Pushmeet
and Riedmiller, Martin
and Vinyals, Oriol
and Silver, David},
title={Faster sorting algorithms discovered using deep reinforcement learning},
journal={Nature},
year={2023},
month={Jun},
day={01},
volume={618},
number={7964},
pages={257-263},
abstract={Fundamental algorithms such as sorting or hashing are used trillions of times on any given day1. As demand for computation grows, it has become critical for these algorithms to be as performant as possible. Whereas remarkable progress has been achieved in the past2, making further improvements on the efficiency of these routines has proved challenging for both human scientists and computational approaches. Here we show how artificial intelligence can go beyond the current state of the art by discovering hitherto unknown routines. To realize this, we formulated the task of finding a better sorting routine as a single-player game. We then trained a new deep reinforcement learning agent, AlphaDev, to play this game. AlphaDev discovered small sorting algorithms from scratch that outperformed previously known human benchmarks. These algorithms have been integrated into the LLVM standard C++ sort library3. This change to this part of the sort library represents the replacement of a component with an algorithm that has been automatically discovered using reinforcement learning. We also present results in extra domains, showcasing the generality of the approach.},
issn={1476-4687},
doi={10.1038/s41586-023-06004-9},
url={https://doi.org/10.1038/s41586-023-06004-9}
}