25,161
edits
Changes
Created page with "'''Home * Engines * Maia Chess''' FILE:Hermes Maia Staatliche Antikensammlungen 2304.jpg|border|right|thumb|220px| Hermes and Maia <ref>God council in [h..."
'''[[Main Page|Home]] * [[Engines]] * Maia Chess'''
[[FILE:Hermes Maia Staatliche Antikensammlungen 2304.jpg|border|right|thumb|220px| Hermes and Maia
<ref>God council in [https://en.wikipedia.org/wiki/Mount_Olympus Olympus]: [https://en.wikipedia.org/wiki/Hermes Hermes] with his mother [https://en.wikipedia.org/wiki/Maia Maia]. Detail of the side B of an [https://en.wikipedia.org/wiki/Attica Attic] [https://en.wikipedia.org/wiki/Red-figure_pottery red-figure] [https://en.wikipedia.org/wiki/Amphora belly-amphora] by [https://en.wikipedia.org/wiki/Nikoxenos_Painter Nikoxenos Painter], ca. 500 BC, [https://en.wikipedia.org/wiki/Staatliche_Antikensammlungen Staatliche Antikensammlungen],
[https://commons.wikimedia.org/wiki/File:Hermes_Maia_Staatliche_Antikensammlungen_2304.jpg image] by [https://commons.wikimedia.org/wiki/User:Bibi_Saint-Pol Bibi Saint-Pol], February 10, 2007, [https://en.wikipedia.org/wiki/Wikimedia_Commons Wikimedia Commons]</ref> ]]
'''Maia Chess''', (Maia, Maiachess)<br/>
a chess engine featuring [[Deep Learning|deep learning]] - as elaborated by their authors [[Reid McIlroy-Young]], [[Siddhartha Sen]], [[Jon Kleinberg]], and [[Ashton Anderson]] in their research paper presented at the 26th [[ACM#SIGKDD|ACM SIGKDD]] virtual conference in 2020 - with the aim to align [https://en.wikipedia.org/wiki/Superhuman superhuman] [[Artificial Intelligence|AI]] with [https://en.wikipedia.org/wiki/Human_behavior human behaviour] <ref> [[Reid McIlroy-Young]], [[Siddhartha Sen]], [[Jon Kleinberg]], [[Ashton Anderson]] ('''2020'''). ''Aligning Superhuman AI with Human Behavior: Chess as a Model System''. In Proceedings of the 26th [[ACM#SIGKDD|ACM SIGKDD 2020]], [https://arxiv.org/abs/2006.01855 arXiv:2006.01855]</ref>.
Like [[AlphaZero]] and [[Leela Chess Zero]], Maia Chess uses a [[Neural Networks#Deep|deep]] [[Neural Networks#Convolutional|convolutional neural network]] (CNN) to predict moves.
In contrast to the '''Zero''' training approaches of their inspirer, using [[Reinforcement Learning|reinforcement learning]] - Maia models are entirely trained by [[Supervised Learning|supervised learning]],
feeding in games of human players separated by 9 [[Playing Strength|rating levels]] between 1100 and 1900 Elo. Further, Maia Chess only predicts moves by probing the net without any [[Search|search]].
Maia chess is [[:Category:Open Source|open source]] released under the terms under the [[Free Software Foundation#GPL|GPL version 3]],
and consists of [[Python]] code relying on the [https://en.wikipedia.org/wiki/Scikit-learn scikit-learn] library, along with various [https://en.wikipedia.org/wiki/Bash_(Unix_shell) bash] scripts.
To play chess, Maia requires its models used in [[Leela Chess Zero|Lc0]] <ref>[https://github.com/CSSLab/lc0_23 GitHub - CSSLab/lc0_23]</ref> similar to any other Leela weights file - in [[UCI]] mode,
''nodes_1'' needs to disable any [[Search|search]] <ref>[https://github.com/CSSLab/maia-chess/blob/master/README.md maia-chess/README.md at master · CSSLab/maia-chess · GitHub]</ref>.
=Model=
[[Board Representation|Boards were represented]] as a 8×8×17 dimensional [[Array|array]] with the 12 channels [[Pieces#PieceTypeCoding|encoding pieces]], 4 channels encoding [[Castling Rights|castling rights]], and one encoding whether the [[Side to move|active player]] is white.
The [[Neural Networks#Residual|residual]] [[Neural Networks#Convolutional|CNN]] has 6 residual blocks with two set of 2D CNNs with 64 channels and a 3×3 kernel <ref>[[Reid McIlroy-Young]], [[Siddhartha Sen]], [[Jon Kleinberg]], [[Ashton Anderson]] ('''2020'''). ''Aligning Superhuman AI with Human Behavior: Chess as a Model System''. [https://arxiv.org/abs/2006.01855 arXiv:2006.01855]. 7 Supplement</ref>.
=Publications=
* [[Reid McIlroy-Young]], [[Siddhartha Sen]], [[Jon Kleinberg]], [[Ashton Anderson]] ('''2020'''). ''Aligning Superhuman AI with Human Behavior: Chess as a Model System''. In Proceedings of the 26th [[ACM#SIGKDD|ACM SIGKDD 2020]], [https://arxiv.org/abs/2006.01855 arXiv:2006.01855]
=Forum Posts=
* [http://www.talkchess.com/forum3/viewtopic.php?f=7&t=75985 Maiachess] by [[Marc-Philippe Huget]], [[CCC]], December 04, 2020
* [http://www.talkchess.com/forum3/viewtopic.php?f=2&t=69257&start=23 Re: Chess for Android and Electronic ChessBoards] by [[Peter Österlund]], [[CCC]], December 06, 2020 » [[Chess for Android]]
=External Links=
==Chess Engine==
* [https://maiachess.com/ Maia Chess]
* [https://github.com/CSSLab/maia-chess GitHub - CSSLab/maia-chess: Human like chess engines]
* [https://csslab.github.io/Maia-Agreement-Visualizer/ Maia Prediction Visualizer]
==Misc==
* [https://en.wikipedia.org/wiki/Maia Maia from Wikipedia]
* [https://en.wikipedia.org/wiki/Maia_(star) Maia (star) from Wikipedia]
* [https://en.wikipedia.org/wiki/Maia_(disambiguation) Maia (disambiguation) from Wikipedia]
=References=
<references />
'''[[Engines|Up one level]]'''
[[Category:Open Source]]
[[Category:GPL]]
[[Category:UCI]]
[[Category:DCNN]]
[[Category:Python]]
[[Category:Astronomy]]
[[Category:Mythology]]
[[Category:Metaphysics]]
[[Category:Chess Suffix]]
[[FILE:Hermes Maia Staatliche Antikensammlungen 2304.jpg|border|right|thumb|220px| Hermes and Maia
<ref>God council in [https://en.wikipedia.org/wiki/Mount_Olympus Olympus]: [https://en.wikipedia.org/wiki/Hermes Hermes] with his mother [https://en.wikipedia.org/wiki/Maia Maia]. Detail of the side B of an [https://en.wikipedia.org/wiki/Attica Attic] [https://en.wikipedia.org/wiki/Red-figure_pottery red-figure] [https://en.wikipedia.org/wiki/Amphora belly-amphora] by [https://en.wikipedia.org/wiki/Nikoxenos_Painter Nikoxenos Painter], ca. 500 BC, [https://en.wikipedia.org/wiki/Staatliche_Antikensammlungen Staatliche Antikensammlungen],
[https://commons.wikimedia.org/wiki/File:Hermes_Maia_Staatliche_Antikensammlungen_2304.jpg image] by [https://commons.wikimedia.org/wiki/User:Bibi_Saint-Pol Bibi Saint-Pol], February 10, 2007, [https://en.wikipedia.org/wiki/Wikimedia_Commons Wikimedia Commons]</ref> ]]
'''Maia Chess''', (Maia, Maiachess)<br/>
a chess engine featuring [[Deep Learning|deep learning]] - as elaborated by their authors [[Reid McIlroy-Young]], [[Siddhartha Sen]], [[Jon Kleinberg]], and [[Ashton Anderson]] in their research paper presented at the 26th [[ACM#SIGKDD|ACM SIGKDD]] virtual conference in 2020 - with the aim to align [https://en.wikipedia.org/wiki/Superhuman superhuman] [[Artificial Intelligence|AI]] with [https://en.wikipedia.org/wiki/Human_behavior human behaviour] <ref> [[Reid McIlroy-Young]], [[Siddhartha Sen]], [[Jon Kleinberg]], [[Ashton Anderson]] ('''2020'''). ''Aligning Superhuman AI with Human Behavior: Chess as a Model System''. In Proceedings of the 26th [[ACM#SIGKDD|ACM SIGKDD 2020]], [https://arxiv.org/abs/2006.01855 arXiv:2006.01855]</ref>.
Like [[AlphaZero]] and [[Leela Chess Zero]], Maia Chess uses a [[Neural Networks#Deep|deep]] [[Neural Networks#Convolutional|convolutional neural network]] (CNN) to predict moves.
In contrast to the '''Zero''' training approaches of their inspirer, using [[Reinforcement Learning|reinforcement learning]] - Maia models are entirely trained by [[Supervised Learning|supervised learning]],
feeding in games of human players separated by 9 [[Playing Strength|rating levels]] between 1100 and 1900 Elo. Further, Maia Chess only predicts moves by probing the net without any [[Search|search]].
Maia chess is [[:Category:Open Source|open source]] released under the terms under the [[Free Software Foundation#GPL|GPL version 3]],
and consists of [[Python]] code relying on the [https://en.wikipedia.org/wiki/Scikit-learn scikit-learn] library, along with various [https://en.wikipedia.org/wiki/Bash_(Unix_shell) bash] scripts.
To play chess, Maia requires its models used in [[Leela Chess Zero|Lc0]] <ref>[https://github.com/CSSLab/lc0_23 GitHub - CSSLab/lc0_23]</ref> similar to any other Leela weights file - in [[UCI]] mode,
''nodes_1'' needs to disable any [[Search|search]] <ref>[https://github.com/CSSLab/maia-chess/blob/master/README.md maia-chess/README.md at master · CSSLab/maia-chess · GitHub]</ref>.
=Model=
[[Board Representation|Boards were represented]] as a 8×8×17 dimensional [[Array|array]] with the 12 channels [[Pieces#PieceTypeCoding|encoding pieces]], 4 channels encoding [[Castling Rights|castling rights]], and one encoding whether the [[Side to move|active player]] is white.
The [[Neural Networks#Residual|residual]] [[Neural Networks#Convolutional|CNN]] has 6 residual blocks with two set of 2D CNNs with 64 channels and a 3×3 kernel <ref>[[Reid McIlroy-Young]], [[Siddhartha Sen]], [[Jon Kleinberg]], [[Ashton Anderson]] ('''2020'''). ''Aligning Superhuman AI with Human Behavior: Chess as a Model System''. [https://arxiv.org/abs/2006.01855 arXiv:2006.01855]. 7 Supplement</ref>.
=Publications=
* [[Reid McIlroy-Young]], [[Siddhartha Sen]], [[Jon Kleinberg]], [[Ashton Anderson]] ('''2020'''). ''Aligning Superhuman AI with Human Behavior: Chess as a Model System''. In Proceedings of the 26th [[ACM#SIGKDD|ACM SIGKDD 2020]], [https://arxiv.org/abs/2006.01855 arXiv:2006.01855]
=Forum Posts=
* [http://www.talkchess.com/forum3/viewtopic.php?f=7&t=75985 Maiachess] by [[Marc-Philippe Huget]], [[CCC]], December 04, 2020
* [http://www.talkchess.com/forum3/viewtopic.php?f=2&t=69257&start=23 Re: Chess for Android and Electronic ChessBoards] by [[Peter Österlund]], [[CCC]], December 06, 2020 » [[Chess for Android]]
=External Links=
==Chess Engine==
* [https://maiachess.com/ Maia Chess]
* [https://github.com/CSSLab/maia-chess GitHub - CSSLab/maia-chess: Human like chess engines]
* [https://csslab.github.io/Maia-Agreement-Visualizer/ Maia Prediction Visualizer]
==Misc==
* [https://en.wikipedia.org/wiki/Maia Maia from Wikipedia]
* [https://en.wikipedia.org/wiki/Maia_(star) Maia (star) from Wikipedia]
* [https://en.wikipedia.org/wiki/Maia_(disambiguation) Maia (disambiguation) from Wikipedia]
=References=
<references />
'''[[Engines|Up one level]]'''
[[Category:Open Source]]
[[Category:GPL]]
[[Category:UCI]]
[[Category:DCNN]]
[[Category:Python]]
[[Category:Astronomy]]
[[Category:Mythology]]
[[Category:Metaphysics]]
[[Category:Chess Suffix]]
