{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,5,8]],"date-time":"2026-05-08T07:09:46Z","timestamp":1778224186852,"version":"3.51.4"},"reference-count":64,"publisher":"Springer Science and Business Media LLC","issue":"7930","license":[{"start":{"date-parts":[[2022,10,5]],"date-time":"2022-10-05T00:00:00Z","timestamp":1664928000000},"content-version":"tdm","delay-in-days":0,"URL":"https:\/\/creativecommons.org\/licenses\/by\/4.0"},{"start":{"date-parts":[[2022,10,5]],"date-time":"2022-10-05T00:00:00Z","timestamp":1664928000000},"content-version":"vor","delay-in-days":0,"URL":"https:\/\/creativecommons.org\/licenses\/by\/4.0"}],"content-domain":{"domain":["link.springer.com"],"crossmark-restriction":false},"short-container-title":["Nature"],"published-print":{"date-parts":[[2022,10,6]]},"abstract":"<jats:title>Abstract<\/jats:title>\n                  <jats:p>\n                    Improving the efficiency of algorithms for fundamental computations can have a widespread impact, as it can affect the overall speed of a large amount of computations. Matrix multiplication is one such primitive task, occurring in many systems\u2014from neural networks to scientific computing routines. The automatic discovery of algorithms using machine learning offers the prospect of reaching beyond human intuition and outperforming the current best human-designed algorithms. However, automating the algorithm discovery procedure is intricate, as the space of possible algorithms is enormous. Here we report a deep reinforcement learning approach based on AlphaZero\n                    <jats:sup>1<\/jats:sup>\n                    for discovering efficient and provably correct algorithms for the multiplication of arbitrary matrices. Our agent, AlphaTensor, is trained to play a single-player game where the objective is finding tensor decompositions within a finite factor space. AlphaTensor discovered algorithms that outperform the state-of-the-art complexity for many matrix sizes. Particularly relevant is the case of 4\u2009\u00d7\u20094 matrices in a finite field, where AlphaTensor\u2019s algorithm improves on Strassen\u2019s two-level algorithm for the first time, to our knowledge, since its discovery 50 years ago\n                    <jats:sup>2<\/jats:sup>\n                    . We further showcase the flexibility of AlphaTensor through different use-cases: algorithms with state-of-the-art complexity for structured matrix multiplication and improved practical efficiency by optimizing matrix multiplication for runtime on specific hardware. Our results highlight AlphaTensor\u2019s ability to accelerate the process of algorithmic discovery on a range of problems, and to optimize for different criteria.\n                  <\/jats:p>","DOI":"10.1038\/s41586-022-05172-4","type":"journal-article","created":{"date-parts":[[2022,10,5]],"date-time":"2022-10-05T12:02:59Z","timestamp":1664971379000},"page":"47-53","update-policy":"https:\/\/doi.org\/10.1007\/springer_crossmark_policy","source":"Crossref","is-referenced-by-count":404,"title":["Discovering faster matrix multiplication algorithms with reinforcement learning"],"prefix":"10.1038","volume":"610","author":[{"ORCID":"https:\/\/orcid.org\/0000-0001-7341-1917","authenticated-orcid":false,"given":"Alhussein","family":"Fawzi","sequence":"first","affiliation":[]},{"given":"Matej","family":"Balog","sequence":"additional","affiliation":[]},{"given":"Aja","family":"Huang","sequence":"additional","affiliation":[]},{"given":"Thomas","family":"Hubert","sequence":"additional","affiliation":[]},{"given":"Bernardino","family":"Romera-Paredes","sequence":"additional","affiliation":[]},{"ORCID":"https:\/\/orcid.org\/0000-0002-8470-8203","authenticated-orcid":false,"given":"Mohammadamin","family":"Barekatain","sequence":"additional","affiliation":[]},{"given":"Alexander","family":"Novikov","sequence":"additional","affiliation":[]},{"given":"Francisco J.","family":"R. Ruiz","sequence":"additional","affiliation":[]},{"given":"Julian","family":"Schrittwieser","sequence":"additional","affiliation":[]},{"given":"Grzegorz","family":"Swirszcz","sequence":"additional","affiliation":[]},{"ORCID":"https:\/\/orcid.org\/0000-0002-5197-2892","authenticated-orcid":false,"given":"David","family":"Silver","sequence":"additional","affiliation":[]},{"ORCID":"https:\/\/orcid.org\/0000-0003-2812-9917","authenticated-orcid":false,"given":"Demis","family":"Hassabis","sequence":"additional","affiliation":[]},{"ORCID":"https:\/\/orcid.org\/0000-0002-7466-7997","authenticated-orcid":false,"given":"Pushmeet","family":"Kohli","sequence":"additional","affiliation":[]}],"member":"297","published-online":{"date-parts":[[2022,10,5]]},"reference":[{"key":"5172_CR1","doi-asserted-by":"publisher","first-page":"1140","DOI":"10.1126\/science.aar6404","volume":"362","author":"D Silver","year":"2018","unstructured":"Silver, D. et al. A general reinforcement learning algorithm that masters chess, shogi, and Go through self-play. Science 362, 1140\u20131144 (2018).","journal-title":"Science"},{"key":"5172_CR2","doi-asserted-by":"publisher","first-page":"354","DOI":"10.1007\/BF02165411","volume":"13","author":"V Strassen","year":"1969","unstructured":"Strassen, V. Gaussian elimination is not optimal. Numer. Math. 13, 354\u2013356 (1969).","journal-title":"Numer. Math."},{"key":"5172_CR3","unstructured":"B\u00fcrgisser, P., Clausen, M. & Shokrollahi, A. Algebraic Complexity Theory Vol. 315 (Springer Science & Business Media, 2013)."},{"key":"5172_CR4","doi-asserted-by":"crossref","unstructured":"Bl\u00e4ser, M. Fast matrix multiplication. Theory Comput. 5, 1\u201360 (2013).","DOI":"10.1145\/2493252.2493254"},{"key":"5172_CR5","doi-asserted-by":"crossref","unstructured":"Landsberg, J. M. Geometry and Complexity Theory 169 (Cambridge Univ. Press, 2017).","DOI":"10.1017\/9781108183192"},{"key":"5172_CR6","unstructured":"Pan, V. Y. Fast feasible and unfeasible matrix multiplication. Preprint at https:\/\/arxiv.org\/abs\/1804.04102 (2018)."},{"key":"5172_CR7","doi-asserted-by":"publisher","first-page":"555","DOI":"10.1017\/S0962492921000076","volume":"30","author":"L-H Lim","year":"2021","unstructured":"Lim, L.-H. Tensors in computations. Acta Numer. 30, 555\u2013764 (2021).","journal-title":"Acta Numer."},{"key":"5172_CR8","doi-asserted-by":"publisher","first-page":"434","DOI":"10.1137\/0210032","volume":"10","author":"A Sch\u00f6nhage","year":"1981","unstructured":"Sch\u00f6nhage, A. Partial and total matrix multiplication. SIAM J. Comput. 10, 434\u2013455 (1981).","journal-title":"SIAM J. Comput."},{"key":"5172_CR9","doi-asserted-by":"crossref","unstructured":"Coppersmith, D. & Winograd, S. Matrix multiplication via arithmetic progressions. In ACM Symposium on Theory of Computing 1\u20136 (ACM, 1987).","DOI":"10.1145\/28395.28396"},{"key":"5172_CR10","doi-asserted-by":"crossref","unstructured":"Strassen, V. The asymptotic spectrum of tensors and the exponent of matrix multiplication. In 27th Annual Symposium on Foundations of Computer Science 49\u201354 (IEEE, 1986).","DOI":"10.1109\/SFCS.1986.52"},{"key":"5172_CR11","doi-asserted-by":"crossref","unstructured":"Le Gall, F. Powers of tensors and fast matrix multiplication. In International Symposium on Symbolic and Algebraic Computation 296\u2013303 (ACM, 2014).","DOI":"10.1145\/2608628.2608664"},{"key":"5172_CR12","doi-asserted-by":"crossref","unstructured":"Alman, J. & Williams, V. V. A refined laser method and faster matrix multiplication. In ACM-SIAM Symposium on Discrete Algorithms 522\u2013539 (SIAM, 2021).","DOI":"10.1137\/1.9781611976465.32"},{"key":"5172_CR13","unstructured":"Gauss, C. F. Theoria Motus Corporum Coelestium in Sectionibus Conicis Solum Ambientium (Perthes and Besser, 1809)."},{"key":"5172_CR14","doi-asserted-by":"publisher","first-page":"1","DOI":"10.1145\/2512329","volume":"60","author":"CJ Hillar","year":"2013","unstructured":"Hillar, C. J. & Lim, L.-H. Most tensor problems are NP-hard. J. ACM 60, 1\u201339 (2013).","journal-title":"J. ACM"},{"key":"5172_CR15","doi-asserted-by":"crossref","unstructured":"Laderman, J. D. A noncommutative algorithm for multiplying 3\u2009\u00d7\u20093 matrices using 23 multiplications. Bull. Am. Math. Soc. 82, 126\u2013128 (1976).","DOI":"10.1090\/S0002-9904-1976-13988-2"},{"key":"5172_CR16","doi-asserted-by":"publisher","first-page":"30","DOI":"10.1137\/0120004","volume":"20","author":"JE Hopcroft","year":"1971","unstructured":"Hopcroft, J. E. & Kerr, L. R. On minimizing the number of multiplications necessary for matrix multiplication. SIAM J. Appl. Math. 20, 30\u201336 (1971).","journal-title":"SIAM J. Appl. Math."},{"key":"5172_CR17","unstructured":"Vervliet, N., Debals, O., Sorber, L., Van Barel, M. & De Lathauwer, L. Tensorlab 3.0 (2016); https:\/\/www.tensorlab.net\/"},{"key":"5172_CR18","doi-asserted-by":"publisher","first-page":"1781","DOI":"10.1134\/S0965542513120129","volume":"53","author":"AV Smirnov","year":"2013","unstructured":"Smirnov, A. V. The bilinear complexity and practical algorithms for matrix multiplication. Comput. Math. Math. Phys. 53, 1781\u20131795 (2013).","journal-title":"Comput. Math. Math. Phys."},{"key":"5172_CR19","doi-asserted-by":"crossref","unstructured":"Sedoglavic, A. & Smirnov, A. V. The tensor rank of 5x5 matrices multiplication is bounded by 98 and its border rank by 89. In Proc. 2021 on International Symposium on Symbolic and Algebraic Computation 345\u2013351 (ACM, 2021).","DOI":"10.1145\/3452143.3465537"},{"key":"5172_CR20","doi-asserted-by":"publisher","first-page":"899","DOI":"10.1016\/j.jsc.2020.10.003","volume":"104","author":"MJ Heule","year":"2021","unstructured":"Heule, M. J., Kauers, M. & Seidl, M. New ways to multiply 3\u2009\u00d7\u20093-matrices. J. Symb. Comput. 104, 899\u2013916 (2021).","journal-title":"J. Symb. Comput."},{"key":"5172_CR21","unstructured":"Hubert, T. et al. Learning and planning in complex action spaces. In International Conference on Machine Learning 4476\u20134486 (PMLR, 2021)."},{"key":"5172_CR22","unstructured":"Zhang, W. & Dietterich, T. G. A reinforcement learning approach to job-shop scheduling. In International Joint Conferences on Artificial Intelligence Vol. 95, 1114\u20131120 (Morgan Kaufmann Publishers, 1995)."},{"key":"5172_CR23","unstructured":"Vaswani, A. Attention is all you need. In International Conference on Neural Information Processing Systems Vol 30, 5998\u20136008 (Curran Associates, 2017)."},{"key":"5172_CR24","unstructured":"Ho, J., Kalchbrenner, N., Weissenborn, D. & Salimans, T. Axial attention in multidimensional transformers. Preprint at https:\/\/arxiv.org\/abs\/1912.12180 (2019)."},{"key":"5172_CR25","doi-asserted-by":"publisher","first-page":"2219","DOI":"10.1016\/j.tcs.2010.12.012","volume":"412","author":"C-\u00c9 Drevet","year":"2011","unstructured":"Drevet, C.-\u00c9., Islam, M. N. & Schost, \u00c9. Optimization techniques for small matrix multiplication. Theor. Comput. Sci. 412, 2219\u20132236 (2011).","journal-title":"Theor. Comput. Sci."},{"key":"5172_CR26","unstructured":"Sedoglavic, A. A non-commutative algorithm for multiplying (7\u2009\u00d7\u20097) matrices using 250 multiplications. Preprint at https:\/\/arxiv.org\/abs\/1712.07935 (2017)."},{"key":"5172_CR27","unstructured":"Battaglia, P. W. et al. Relational inductive biases, deep learning, and graph networks. Preprint at https:\/\/arxiv.org\/abs\/1806.01261 (2018)."},{"key":"5172_CR28","unstructured":"Balog, M., van Merri\u00ebnboer, B., Moitra, S., Li, Y. & Tarlow, D. Fast training of sparse graph neural networks on dense hardware. Preprint at https:\/\/arxiv.org\/abs\/1906.11786 (2019)."},{"key":"5172_CR29","doi-asserted-by":"publisher","first-page":"45","DOI":"10.1007\/s10208-016-9332-x","volume":"18","author":"K Ye","year":"2018","unstructured":"Ye, K. & Lim, L.-H. Fast structured matrix computations: tensor rank and Cohn\u2013Umans method. Found. Comput. Math. 18, 45\u201395 (2018).","journal-title":"Found. Comput. Math."},{"key":"5172_CR30","unstructured":"Bradbury, J. et al. JAX: composable transformations of Python+NumPy programs. GitHub http:\/\/github.com\/google\/jax (2018)."},{"key":"5172_CR31","doi-asserted-by":"publisher","first-page":"42","DOI":"10.1145\/2858788.2688513","volume":"50","author":"AR Benson","year":"2015","unstructured":"Benson, A. R. & Ballard, G. A framework for practical parallel fast matrix multiplication. ACM SIGPLAN Not. 50, 42\u201353 (2015).","journal-title":"ACM SIGPLAN Not."},{"key":"5172_CR32","doi-asserted-by":"crossref","unstructured":"Huang, J., Smith, T. M., Henry, G. M. & Van De Geijn, R. A. Strassen\u2019s algorithm reloaded. In International Conference for High Performance Computing, Networking, Storage and Analysis 690\u2013701 (IEEE, 2016).","DOI":"10.1109\/SC.2016.58"},{"key":"5172_CR33","unstructured":"Abadi, M. et al. Tensorflow: a system for large-scale machine learning. In USENIX Symposium On Operating Systems Design And Implementation 265\u2013283 (USENIX, 2016)."},{"key":"5172_CR34","doi-asserted-by":"crossref","unstructured":"Dabney, W., Rowland, M., Bellemare, M. & Munos, R. Distributional reinforcement learning with quantile regression. In AAAI Conference on Artificial Intelligence Vol. 32, 2892\u20132901 (AAAI Press, 2018).","DOI":"10.1609\/aaai.v32i1.11791"},{"key":"5172_CR35","unstructured":"Kingma, D. P., & Ba, J. Adam: a method for stochastic optimization. In International Conference on Learning Representations (ICLR) (2015)."},{"key":"5172_CR36","unstructured":"Loshchilov, I. & Hutter, F. Decoupled weight decay regularization. In International Conference on Learning Representations (ICLR) (2019)."},{"key":"5172_CR37","doi-asserted-by":"publisher","first-page":"484","DOI":"10.1038\/nature16961","volume":"529","author":"D Silver","year":"2016","unstructured":"Silver, D. et al. Mastering the game of Go with deep neural networks and tree search. Nature 529, 484\u2013489 (2016).","journal-title":"Nature"},{"key":"5172_CR38","unstructured":"Sedoglavic, A. A non-commutative algorithm for multiplying 5x5 matrices using 99 multiplications. Preprint at https:\/\/arxiv.org\/abs\/1707.06860 (2017)."},{"key":"5172_CR39","doi-asserted-by":"publisher","first-page":"127","DOI":"10.1016\/0304-3975(78)90045-2","volume":"7","author":"HF de Groote","year":"1978","unstructured":"de Groote, H. F. On varieties of optimal algorithms for the computation of bilinear mappings II. optimal algorithms for 2\u2009\u00d7\u20092-matrix multiplication. Theor. Comput. Sci. 7, 127\u2013148 (1978).","journal-title":"Theor. Comput. Sci."},{"key":"5172_CR40","unstructured":"Burichenko, V. P. On symmetries of the Strassen algorithm. Preprint at https:\/\/arxiv.org\/abs\/1408.6273 (2014)."},{"key":"5172_CR41","doi-asserted-by":"publisher","first-page":"322","DOI":"10.1080\/10586458.2017.1403981","volume":"28","author":"L Chiantini","year":"2019","unstructured":"Chiantini, L., Ikenmeyer, C., Landsberg, J. M. & Ottaviani, G. The geometry of rank decompositions of matrix multiplication I: 2\u2009\u00d7\u20092 matrices. Exp. Math. 28, 322\u2013327 (2019).","journal-title":"Exp. Math."},{"key":"5172_CR42","unstructured":"Grochow, J. A. & Moore, C. Designing Strassen\u2019s algorithm. Preprint at https:\/\/arxiv.org\/abs\/1708.09398 (2017)."},{"key":"5172_CR43","doi-asserted-by":"publisher","first-page":"3205","DOI":"10.1016\/j.jpaa.2018.10.014","volume":"223","author":"G Ballard","year":"2019","unstructured":"Ballard, G., Ikenmeyer, C., Landsberg, J. M. & Ryder, N. The geometry of rank decompositions of matrix multiplication II: 3\u2009\u00d7\u20093 matrices. J. Pure Appl. Algebra 223, 3205\u20133224 (2019).","journal-title":"J. Pure Appl. Algebra"},{"key":"5172_CR44","unstructured":"Grochow, J. A. & Moore, C. Matrix multiplication algorithms from group orbits. Preprint at https:\/\/arxiv.org\/abs\/1612.01527 (2016)."},{"key":"5172_CR45","doi-asserted-by":"publisher","first-page":"455","DOI":"10.1137\/07070111X","volume":"51","author":"TG Kolda","year":"2009","unstructured":"Kolda, T. G. & Bader, B. W. Tensor decompositions and applications. SIAM Rev. 51, 455\u2013500 (2009).","journal-title":"SIAM Rev."},{"key":"5172_CR46","doi-asserted-by":"publisher","first-page":"51","DOI":"10.1016\/j.jsc.2012.05.012","volume":"52","author":"A Bernardi","year":"2013","unstructured":"Bernardi, A., Brachat, J., Comon, P. & Mourrain, B. General tensor decomposition, moment matrices and applications. J. Symb. Comput. 52, 51\u201371 (2013).","journal-title":"J. Symb. Comput."},{"key":"5172_CR47","first-page":"3964","volume":"17","author":"V Elser","year":"2016","unstructured":"Elser, V. A network that learns Strassen multiplication. J. Mach. Learn. Res. 17, 3964\u20133976 (2016).","journal-title":"J. Mach. Learn. Res."},{"key":"5172_CR48","unstructured":"Tschannen, M., Khanna, A. & Anandkumar, A, StrassenNets: deep learning with a multiplication budget. In International Conference on Machine Learning 4985\u20134994 (PMLR, 2018)."},{"key":"5172_CR49","doi-asserted-by":"publisher","first-page":"1","DOI":"10.1145\/3372419","volume":"46","author":"J Huang","year":"2020","unstructured":"Huang, J., Yu, C. D. & Geijn, R. A. V. D. Strassen\u2019s algorithm reloaded on GPUs. ACM Trans. Math. Softw. 46, 1\u201322 (2020).","journal-title":"ACM Trans. Math. Softw."},{"key":"5172_CR50","doi-asserted-by":"publisher","first-page":"207","DOI":"10.1038\/s41586-021-03544-w","volume":"594","author":"A Mirhoseini","year":"2021","unstructured":"Mirhoseini, A. et al. A graph placement methodology for fast chip design. Nature 594, 207\u2013212 (2021).","journal-title":"Nature"},{"key":"5172_CR51","unstructured":"Bunel, R., Desmaison, A., Kohli, P., Torr, P. H. & Kumar, M. P. Learning to superoptimize programs. In International Conference on Learning Representations (ICLR) (2017)."},{"key":"5172_CR52","unstructured":"Li, Y., Gimeno, F., Kohli, P. & Vinyals, O. Strong generalization and efficiency in neural programs. Preprint at https:\/\/arxiv.org\/abs\/2007.03629 (2020)."},{"key":"5172_CR53","unstructured":"Lagoudakis, M. G. et al. Algorithm selection using reinforcement learning. In International Conference on Machine Learning 511\u2013518 (Morgan Kaufmann Publishers, 2000)."},{"key":"5172_CR54","unstructured":"Schmidhuber, J. Evolutionary Principles in Self-Referential Learning. On Learning now to Learn: The Meta-Meta-Meta...-Hook. Diploma thesis, Technische Univ. Munchen (1987)."},{"key":"5172_CR55","unstructured":"Kaliszyk, C., Urban, J., Michalewski, H. & Ol\u0161\u00e1k, M. Reinforcement learning of theorem proving. In International Conference on Neural Information Processing Systems 8836\u20138847 (Curran Associates, 2018)."},{"key":"5172_CR56","doi-asserted-by":"crossref","unstructured":"Piotrowski, B. & Urban, J. ATPboost: learning premise selection in binary setting with ATP feedback. In International Joint Conference on Automated Reasoning 566\u2013574 (Springer, 2018).","DOI":"10.1007\/978-3-319-94205-6_37"},{"key":"5172_CR57","unstructured":"Bansal, K., Loos, S., Rabe, M., Szegedy, C. & Wilcox, S. HOList: an environment for machine learning of higher order logic theorem proving. In International Conference on Machine Learning 454\u2013463 (PMLR, 2019)."},{"key":"5172_CR58","doi-asserted-by":"crossref","unstructured":"Zombori, Z., Urban, J. & Brown, C. E. Prolog technology reinforcement learning prover. In International Joint Conference on Automated Reasoning 489\u2013507 (Springer, 2020).","DOI":"10.1007\/978-3-030-51054-1_33"},{"key":"5172_CR59","unstructured":"Wagner, A. Z. Constructions in combinatorics via neural networks. Preprint at https:\/\/arxiv.org\/abs\/2104.14516 (2021)."},{"key":"5172_CR60","doi-asserted-by":"crossref","unstructured":"Popova, M., Isayev, O. & Tropsha, A. Deep reinforcement learning for de novo drug design. Sci. Adv. 4, eaap7885 (2018).","DOI":"10.1126\/sciadv.aap7885"},{"key":"5172_CR61","doi-asserted-by":"publisher","first-page":"10752","DOI":"10.1038\/s41598-019-47148-x","volume":"9","author":"Z Zhou","year":"2019","unstructured":"Zhou, Z., Kearnes, S., Li, L., Zare, R. N. & Riley, P. Optimization of molecules via deep reinforcement learning. Sci. Rep. 9, 10752 (2019).","journal-title":"Sci. Rep."},{"key":"5172_CR62","doi-asserted-by":"publisher","first-page":"604","DOI":"10.1038\/nature25978","volume":"555","author":"MH Segler","year":"2018","unstructured":"Segler, M. H., Preuss, M. & Waller, M. P. Planning chemical syntheses with deep neural networks and symbolic AI. Nature 555, 604\u2013610 (2018).","journal-title":"Nature"},{"key":"5172_CR63","doi-asserted-by":"publisher","first-page":"6","DOI":"10.1038\/s41534-019-0241-0","volume":"6","author":"M Dalgaard","year":"2020","unstructured":"Dalgaard, M., Motzoi, F., S\u00f8rensen, J. J. & Sherson, J. Global optimization of quantum dynamics with AlphaZero deep exploration. npj Quantum Inf. 6, 6 (2020).","journal-title":"npj Quantum Inf."},{"key":"5172_CR64","unstructured":"Fast matrix multiplication algorithms catalogue. Universit\u00e9 de Lille https:\/\/fmm.univ-lille.fr\/ (2021)."}],"container-title":["Nature"],"original-title":[],"language":"en","link":[{"URL":"https:\/\/www.nature.com\/articles\/s41586-022-05172-4.pdf","content-type":"application\/pdf","content-version":"vor","intended-application":"text-mining"},{"URL":"https:\/\/www.nature.com\/articles\/s41586-022-05172-4","content-type":"text\/html","content-version":"vor","intended-application":"text-mining"},{"URL":"https:\/\/www.nature.com\/articles\/s41586-022-05172-4.pdf","content-type":"application\/pdf","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2022,10,5]],"date-time":"2022-10-05T19:04:42Z","timestamp":1664996682000},"score":1,"resource":{"primary":{"URL":"https:\/\/www.nature.com\/articles\/s41586-022-05172-4"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2022,10,5]]},"references-count":64,"journal-issue":{"issue":"7930","published-print":{"date-parts":[[2022,10,6]]}},"alternative-id":["5172"],"URL":"https:\/\/doi.org\/10.1038\/s41586-022-05172-4","relation":{"has-review":[{"id-type":"doi","id":"10.3410\/f.742353014.793595962","asserted-by":"object"}]},"ISSN":["0028-0836","1476-4687"],"issn-type":[{"value":"0028-0836","type":"print"},{"value":"1476-4687","type":"electronic"}],"subject":[],"published":{"date-parts":[[2022,10,5]]},"assertion":[{"value":"2 October 2021","order":1,"name":"received","label":"Received","group":{"name":"ArticleHistory","label":"Article History"}},{"value":"2 August 2022","order":2,"name":"accepted","label":"Accepted","group":{"name":"ArticleHistory","label":"Article History"}},{"value":"5 October 2022","order":3,"name":"first_online","label":"First Online","group":{"name":"ArticleHistory","label":"Article History"}},{"value":"The authors of the paper are planning to file a patent application relating to subject matter contained in this paper in the name of DeepMind Technologies Limited.","order":1,"name":"Ethics","group":{"name":"EthicsHeading","label":"Competing interests"}}]}}