The DOI matches its Crossref record; 30 of the 30 references it cites were found in a registry or an index; 5 of the 16 sources its claims cite were checked for support.
A general reinforcement learning algorithm that masters chess, shogi, and Go through self-play
Open accessAbstract
One program to rule them all: Computers can beat humans at increasingly complex games, including chess and Go. However, these programs are typically constructed for a particular game, exploiting its properties, such as the symmetries of the board on which it is played. Silver et al. developed a program called AlphaZero, which taught itself to play Go, chess, and shogi (a Japanese version of chess) (see the Editorial, and the Perspective by Campbell). AlphaZero managed to beat state-of-the-art programs specializing in these three games. The ability of AlphaZero to adapt to various game rules is a notable step toward achieving a general game-playing system. Science , this issue p. 1140 ; see also pp. 1087 and 1118
Overview
CoChat Summary
TL;DR
AlphaZero is a single reinforcement-learning and MCTS algorithm that, given only game rules, learned superhuman play in chess, shogi, and Go through self-play. It defeated world-champion programs Stockfish, Elmo, and AlphaGo Zero using the same algorithm and network architecture across all three games.
Key findings
- In chess, AlphaZero defeated Stockfish 8, winning 155 games and losing 6 out of 1,000; it also won matches from common human openings and 2016 TCEC opening positions.
- In shogi, AlphaZero defeated Elmo, winning 98.2% of games as black and 91.2% overall; in Go, it defeated AlphaGo Zero, winning 61% of games.
- AlphaZero first outperformed Stockfish after 4 hours (300,000 steps), Elmo after 2 hours (110,000 steps), and AlphaGo Lee after 30 hours (74,000 steps); training ran for 700,000 steps, taking about 9 hours in chess, 12 hours in shogi, and 13 days in Go.
- AlphaZero searched only about 60,000 positions per second in chess and shogi, versus 60 million for Stockfish and 25 million for Elmo, yet still defeated Stockfish with 1/10 thinking time and won 46% against Elmo with 1/100 thinking time.
Methods
- AlphaZero uses a deep neural network \(f_\theta(s)\) that outputs move probabilities \(p\) and a scalar value \(v\), trained by reinforcement learning from self-play starting from random parameters; MCTS uses these outputs to guide search, with 800 simulations per move during training.
- The same algorithm, network architecture, and hyperparameters were used for chess, shogi, and Go, with no domain knowledge except game rules; no opening book, endgame tablebases, or domain-specific heuristics were used.
- Training used 5,000 first-generation TPUs to generate self-play games and 16 second-generation TPUs to train networks; evaluation matches used 3 hours per game plus 15 seconds per move, with opponents on 44 CPU cores and AlphaZero on four first-generation TPUs and 44 CPU cores.
Limitations
- Stated: AlphaZero did not use opening books, endgame tablebases, or domain-specific heuristics; the authors note that some such techniques could further improve performance but were left for future research.
- Stated: hyperparameters were reused across games without game-specific tuning, except for exploration noise and the learning rate schedule; matches against an opening book were low in diversity, and forced randomization increased losses.
- Evident from the text: results cover only three games with separate trained instances, no cross-game transfer is demonstrated, and evaluation compares AlphaZero on TPU hardware against CPU-based opponents, so hardware and time-control differences qualify direct strength comparisons.
Model-written summary. Check it against the paper before relying on it.
Integrity signals
Registry
Cites
Statistics
Verification
Claims the paper makes
3 Valid · 2 Not valid · 8 Not checked · 43 Own results and interpretations
Every source the paper cites for this claim was checked and supports it.
In the paper
“Apart from the representation of positions and actions described above, AlphaZero uses the same network architecture as AlphaGo Zero ( 9 ), briefly recapitulated here.”
What the sources say
- Supportedcites [9] source
The Neural Network Architecture Comparison (p. 29) identifies AlphaGo Zero’s architecture as a residual tower with policy and value heads. The citation supplies the AlphaGo Zero architecture that the citing authors say AlphaZero uses; the comparison to AlphaZero is their own statement. (LLM, conf=0.88)
plain- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: unclear — It is unclear whether the paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
- Supported
The cited sources support only part of this claim — read each source’s verdict below.
In the paper
“and in Go, AlphaZero first outperformed AlphaGo Lee ( 9 ) after 30 hours (74,000 steps).”
What the sources say
- Partly supportedcites [9] sourceExperimental — accuracy not yet measured
Section 2, p. 7 reports that AlphaGo Zero outperformed AlphaGo Lee, but says this happened after 36 hours, not 30. The passage does not give 74,000 steps, and it names the system AlphaGo Zero rather than AlphaZero. (LLM, conf=0.90)
plain- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: unclear — It is unclear whether the paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
- Partly supported
A result the paper reports from its own data — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“The training algorithm achieved similar performance in all independent runs (see fig. S3), suggesting that the high performance of AlphaZero's training algorithm is repeatable.”
plain- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: passes — The paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
A result the paper reports from its own data — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“AlphaZero also defeated Stockfish when given 1 / 10 as much thinking time as its opponent (i.e. searching ~ 1 / 10 , 000 as many positions),”
plainstates a cause- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: passes — The paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
A result the paper reports from its own data — it cites no other source for it, so there is none to check. The quote shows it is in the paper.
In the paper
“and won 46% of games against Elmo when given 1 / 100 as much time (i.e. searching ~ 1 / 40 , 000 as many positions), see Fig. 2.”
plain- Asserted: passes — The authors put it forward as true.
- Checkable: passes — It can be checked against evidence.
- Load-bearing: passes — The paper’s conclusions depend on it.
- Based: passes — The paper gives a reason why it holds.
23 more claims
Create a free account to see all 56 claims and to verify the 3 references still open.
- Probably not a substantial claim. 10
- It may state two separate claims as one. 23
- The quote only partly states this claim. 1
Claims other papers make about this one
Each is checked against this paper’s own text: does it say what the citing paper says it does?
The source backs part of the claim, not all of it — or backs it for a different population, condition or comparison.
In the citing paper
“Games have historically served as highly effective metrics for evaluating progress in deep reinforcement learning research”
From BALROG: Benchmarking Agentic LLM and VLM Reasoning On Games (2024) · cites this paper as Silver et al., 2018
Experimental — accuracy not yet measuredThe source backs part of the claim, not all of it — or backs it for a different population, condition or comparison.
In the citing paper
“In 2017, AlphaGo, using the same learning algorithms in TD-Gammon and a billion times more computing power, soundly beat Ke Jie, the world Go champion, in a game that is to chess what chess is to backgammon in complexity (Silver et al., 2018).”
From Large Language Models and the Reverse Turing Test (2023) · cites this paper as Silver et al., 2018
Experimental — accuracy not yet measuredThe source backs part of the claim, not all of it — or backs it for a different population, condition or comparison.
In the citing paper
“Most of the attention was on who won, but more important was how it won: AlphaGo was creative and had discovered new moves and board positions that were outside human experience.”
From Large Language Models and the Reverse Turing Test (2023) · cites this paper as Silver et al., 2018
Experimental — accuracy not yet measuredThe source backs part of the claim, not all of it — or backs it for a different population, condition or comparison.
In the citing paper
“The primary method for applying more computation at inference, common in the AlphaZero family of agents2,72, is scaling the tree search and running multiple independent search attempts.”
From Olympiad-level formal mathematical reasoning with reinforcement learning (2026) · cites this paper as [2]
Experimental — accuracy not yet measuredThe cited source says what the claim attributes to it.
In the citing paper
“Modern AI can write computer programs (Li et al., 2022) or play chess at super-human level (Silver et al., 2018), but the ability of AI to perceive and act in the world remains far below human level.”
From Scaling Instructable Agents Across Many Simulated Worlds (2024) · cites this paper as Silver et al., 2018
1 more claim about this paper
Create a free account to see every claim other papers make about this one.
Citations
Cites
Mastering the game of Go with deep neural networks and tree search
Silver, D., Huang, A., Maddison, C. J., Guez, A., Sifre, L., van den Driessche, G., Schrittwieser, J., Antonoglou, I., Panneershelvam, V., Lanctot, M., et al. (2016) · Nature · doi:10.1038/nature16961 · Cited by 15,612
Self-citationAdvances in neural information processing systems 7
(1996) · Computers & Mathematics with Applications · doi:10.1016/s0898-1221(96)90463-0 · Cited by 14,281
Identity mappings in deep residual networks
He, K., Zhang, X., Ren, S., & Sun, J. (2016) · Lecture Notes in Computer Science · doi:10.1007/978-3-319-46493-0_38 · Cited by 10,741
Identity Mappings in Deep Residual Networks
He, K., Zhang, X., Ren, S., & Sun, J. (2016) · arXiv · doi:10.48550/arxiv.1603.05027 · Cited by 10,665
Mastering the game of Go without human knowledge
Silver, D., Schrittwieser, J., Simonyan, K., Antonoglou, I., Huang, A., Guez, A., Hubert, T., Baker, L., Lai, M., Bolton, A., et al. (2017) · Nature · doi:10.1038/nature24270 · Cited by 9,084
Self-citationIn-Datacenter Performance Analysis of a Tensor Processing Unit
Jouppi, N. P., Young, C., Patil, N., Patterson, D., Agrawal, G., Bajwa, R., Bates, S., Bhatia, S., Boden, N., Borchers, A., et al. (2017) · Proceedings of the 44th Annual International Symposium on Computer Architecture · doi:10.1145/3079856.3080246 · Cited by 4,537
Encyclopedia of Artificial Intelligence
(2009) · IGI Global eBooks · doi:10.4018/978-1-59904-849-9 · Cited by 1,396
Campbell, M., Hoane, A. J., & Hsu, F.-H. (2002) · Artificial Intelligence · doi:10.1016/s0004-3702(01)00129-1 · Cited by 1,226
Lim, T.-S., Loh, W.-Y., & Shih, Y.-S. (2000) · Machine Learning · doi:10.1023/a:1007608224229 · Cited by 1,133
An analysis of alpha-beta pruning
Knuth, D. E., & Moore, R. W. (1975) · Artificial Intelligence · doi:10.1016/0004-3702(75)90019-3 · Cited by 1,004
Research works printed but not linked: 7
- C. J. Maddison A. Huang I. Sutskever D. Silver paper presented at the International Conference on Learning Representations 2015 San Diego CA 7 to 9 May 2015.
- D. N. L. Levy M. Newborn How Computers Play Chess (Ishi Press 2009).
- W. Steinitz The Modern Chess Instructor (Edition Olms 1990).
- O. Arenz “Monte Carlo chess ” master’s thesis Technische Universität Darmstadt (2012).
- 10.1023/a:1007634325138 · DOI
- T. Kaneko K. Hoki in Advances in Computer Games – 13th International Conference ACG 2011 Revised Selected Papers Tilburg Netherlands 20 to 22 November 2011 (Springer 2012) pp. 158–169.
- R. Ramanujan A. Sabharwal B. Selman in Proceedings of the 26th Conference on Uncertainty in Artificial Intelligence (UAI 2010) Catalina Island CA 8 to 11 July (AUAI Press 2010).
Cited by
Items 901–1,000 of 3,622
Formal Methods for Autonomous Systems
Wongpiromsarn, T., Ghasemi, M., Cubuktepe, M., Bakirtzis, G., Carr, S., Karabag, M. O., Neary, C., Gohari, P., & Topcu, U. (2023) · Foundations and Trends® in Systems and Control · doi:10.1561/2600000029 · Cited by 17
Esports Training, Periodization, and Software—A Scoping Review
Białecki, A., Michalak, B., & Gajewski, J. (2024) · Applied Sciences · doi:10.3390/app142210354 · Cited by 26
“These systems are capable of beating human opponents in drone racing [7], Go [89], chess and shogi [90, 91], StarCraft 2 [92], Dota 2 [93, 94].”
BackgroundLin, X., Chen, W., Zhou, Z., Li, J., Zhao, Y., & Zhang, X. (2025) · Robotics and Computer-Integrated Manufacturing · doi:10.1016/j.rcim.2025.103009 · Cited by 21
Khajehnejad, M. A., Habibollahi, F., Loeffler, A., Paul, A., Razi, A., & Kagan, B. J. (2025) · Cyborg and Bionic Systems · doi:10.34133/cbsystems.0336 · Cited by 37
“Reinforcement algorithms that incorporate deep neural networks have been developed to beat human experts in multiple game settings including poker [ 7 ], multiplayer contests [ 8 ], complex board games including go and chess [ 9 – 11 ], and numerous Atari video games [ 12 ].”
SAI a Sensible Artificial Intelligence that plays Go
Morandin, F., Amato, G., Gini, R., Metta, C., Parton, M., & Pascutto, G.-C. (2019) · International Joint Conference on Neural Networks (IJCNN) · doi:10.1109/ijcnn.2019.8852266 · Cited by 23
Cheng, K. S., Pan, R., Pan, H., Li, B., Meena, S. S., Xing, H., Ng, Y. J., Qin, K., Liao, X., Kosgei, B. K., et al. (2020) · Theranostics · doi:10.7150/thno.44053 · Cited by 17
“Taking hints from Google DeepMind gaming program, AlphaZero [46] that combined deep neural networks and traditional symbolic Monte Carlo tree search, we adopted a similar hybrid AI strategy in ALICE by integrating machine learning with rule-based algorithms to accurately enumerate various CTC…”
MethodTectonic discrimination and application based on convolution neural network and incomplete big data
Ge, C., Huo, J., Gu, H., Wang, F., Sun, H., Li, X., Li, W., & Feng, Y. (2020) · Journal of Geochemical Exploration · doi:10.1016/j.gexplo.2020.106662 · Cited by 18
Behavior Self-Organization Supports Task Inference for Continual Robot Learning
Hafez, M. B., & Wermter, S. (2021) · IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS) · doi:10.1109/iros51168.2021.9636297 · Cited by 18
“Recent advances in reinforcement learning (RL) have shown great success, achieving superhuman performance in a variety of challenging board and Atari games [1], [2] and enabling the acquisition of complex robotic manipulation skills [3], [4] through trial and error without any prior knowledge”
BackgroundInnate heuristics and fast learning support escape route selection in mice
Claudi, F., Campagner, D., & Branco, T. (2022) · Current Biology · doi:10.1016/j.cub.2022.05.020 · Cited by 13
Soranzo, E., Guardiani, C., Saif, A., & Wu, W. (2022) · Computers & Geosciences · doi:10.1016/j.cageo.2022.105182 · Cited by 17
MASAC-based confrontation game method of UAV clusters
王, 尔., 刘, 帆., Hong, C., 郭, 靖., 何, 宁., 赵, 琳., & 薛, 健. (2022) · Scientia Sinica Informationis · doi:10.1360/ssi-2022-0303 · Cited by 16
The paradoxical transparency of opaque machine learning
Lo, F. T. H. (2022) · AI & Society · doi:10.1007/s00146-022-01616-7 · Cited by 20
“It seems awfully wasteful not to adopt the technique of opaque machine learning over the issue of opacity when these algorithms appear to outperform human decision-making in numerous contexts (Longoni et al. 2019; McKin-ney et al. 2020; Silver et al. 2018; Stokes et al. 2020).”
Kuru, K., Clough, S., Ansell, D., McCarthy, J., & McGovern, S. (2023) · Expert Systems with Applications · doi:10.1016/j.eswa.2023.120574 · Cited by 24
“…of Deep Reinforcement Learning (DRL) in numerous fields such as Atari games (Mnih et al., 2015), poker (Moravčík et al., 2017), multiplayer games (Jaderberg et al., 2019), and board games (Silver et al., 2016; Silver, Hubert et al., 2017; Silver et al., 2018; Silver, Schrittwieser et al., 2017).”
BackgroundReconfigurable nonlinear optical element using tunable couplers and inverse‐designed structure
Nikkhah, V., Mencagli, M. J., & Engheta, N. (2023) · Nanophotonics · doi:10.1515/nanoph-2023-0152 · Cited by 16
“These features make them appealing for several applications that require handling large data sets such as real-time image processing [7], language translation [8], decision-making problems [9], and more [10, 11].”
An Adaptive Threshold for the Canny Edge With Actor-Critic Algorithm
Choi, K.-H., & Ha, J.-E. (2023) · IEEE Access · doi:10.1109/access.2023.3291593 · Cited by 19
“[36], object localization [37], region proposal [38], and visual”
Backgroundde Carvalho, J. P., & Dimitrakopoulos, R. G. (2023) · Applied Intelligence · doi:10.1007/s10489-023-04774-3 · Cited by 22
Calibrating machine behavior: a challenge for AI alignment
Firt, E. (2023) · Ethics and Information Technology · doi:10.1007/s10676-023-09716-8 · Cited by 19
“Why not just let a trained and tested system, aligned with human values, infer the best output (in this case, behavior or action of some sort) for a given input (in this case, a certain state of affairs, understood as requiring an action)? After all, we know that the current-days state of the art AI systems can perfor…”
BackgroundBhattacharya, S., Kailas, S., Badyal, S., Gil, S., & Bertsekas, D. P. (2023) · IEEE Transactions on Robotics · doi:10.1109/tro.2023.3347128 · Cited by 34
“AlphaZero [25] discusses a scheme whereby MCTS-based online lookahead is used to improve offline approximation architectures given by deep neural networks in both policy and value spaces for perfect-state observation cases.”
MethodImproving traffic light systems using Deep Q-networks
Moreno, J., Posadas-Yagüe, J.-L., Cano, J., Calafate, C. T., Conejero, J. A., & Poza-Luján, J.-L. (2024) · Expert Systems with Applications · doi:10.1016/j.eswa.2024.124178 · Cited by 31
2-level reinforcement learning for ships on inland waterways: Path planning and following
Waltz, M., Paulig, N., & Okhrin, O. (2025) · Expert Systems with Applications · doi:10.1016/j.eswa.2025.126933 · Cited by 27
Saarbrücken, J. (. 2., Francesconi, E., Borges, G., & Sorge, C. (2022) · Frontiers in artificial intelligence and applications · doi:10.3233/faia362 · Cited by 16
Accelerating and Improving AlphaZero Using Population Based Training
Wu, T.-R., Wei, T.-H., & Wu, I. (2020) · Proceedings of the AAAI Conference on Artificial Intelligence · doi:10.1609/aaai.v34i01.5454 · Cited by 16
“Since AlphaGo successfully defeated the top human Go player in 2016 (Silver et al. 2016), DeepMind followed up with the algorithms AlphaGo Zero (Silver et al. 2017) and AlphaZero (Silver et al. 2018).”
BackgroundMethodInfluentialLibera, A. D., Romeres, D., Jha, D. K., et al. (2020) · IEEE Robotics and Automation Letters · doi:10.1109/lra.2020.2977255 · Cited by 16
“RL algorithms have been able to reach and exceed human-level performance in several benchmark problems, such as playing the games of chess, go and shogi [1].”
BackgroundHwang, J., Tanaka, Y., Ishino, S., & Watanabe, S. (2020) · Science and Technology of Advanced Materials · doi:10.1080/14686996.2020.1786856 · Cited by 17
Deep learning for general game playing with Ludii and Polygames
Soemers, D. J. N. J., Mella, V., Browne, C., & Teytaud, O. (2022) · ICGA Journal · doi:10.3233/icg-220197 · Cited by 16
“Self-play training approaches such as those popularised by AlphaGo Zero (Silver et al., 2017) and AlphaZero (Silver et al., 2018), based on combinations of Monte-Carlo tree search (MCTS) (Kocsis and Szepesvári, 2006; Coulom, 2007; Browne et al., 2012) and Deep Learning (LeCun et al., 2015), have been demonstrated to b…”
InfluentialPraeger, M. F., Xie, Y., Grant‐Jacob, J. A., Eason, R. W., & Mills, B. (2021) · Machine Learning Science and Technology · doi:10.1088/2632-2153/abf0f6 · Cited by 17
Chess as a Testbed for Language Model State Tracking
Toshniwal, S., Wiseman, S., Livescu, K., & Gimpel, K. (2022) · Proceedings of the AAAI Conference on Artificial Intelligence · doi:10.1609/aaai.v36i10.21390 · Cited by 9
“For example, using only self-play and the rules of chess, AlphaZero achieves superhuman performance starting from random play (Silver et al. 2018).”
MethodA Unifying Framework for Reinforcement Learning and Planning
Moerland, T. M., Broekens, J., Plaat, A., & Jonker, C. M. (2022) · Frontiers in Artificial Intelligence · doi:10.3389/frai.2022.908353 · Cited by 45
“…et al., 2002), the first computer programme to defeat a human Chess world champion, used a heuristic bootstrap function, this approach was later outperformed by AlphaZero (Silver et al., 2018), which uses a deep neural network to learn a bootstrap function that provides better generalization.”
BackgroundMethodInfluentialModern Artificial Neural Networks: Is Evolution Cleverer?
Bahmer, A., Gupta, D. S., & Effenberger, F. (2023) · Neural Computation · doi:10.1162/neco_a_01575 · Cited by 19
“Hassabis, D. (2017). Mastering the game of Go without human knowledge.”
BackgroundToward evolutionary and developmental intelligence
Doya, K., & Taniguchi, T. (2019) · Current Opinion in Behavioral Sciences · doi:10.1016/j.cobeha.2019.04.006 · Cited by 21
Learning and Planning in Complex Action SpacesPreprint
Hubert, T., Schrittwieser, J., Antonoglou, I., Barekatain, M., Schmitt, S., & Silver, D. (2021) · arXiv (Cornell University) · doi:10.48550/arxiv.2104.06303 · Cited by 114
“Simultaneously, planning based methods have enjoyed huge successes in domains with discrete action spaces, surpassing human performance in the classical games of chess and Go (Silver et al., 2018) or poker (Brown & Sandholm, 2018; Morav ˇ c ´ ık et al., 2017).”
BackgroundMethodSelf-citationDeco, G., Perl, Y. S., Sitt, J. D., Tagliazucchi, E., & Kringelbach, M. L. (2021) · bioRxiv (Cold Spring Harbor Laboratory) · doi:10.1101/2021.07.02.450899 · Cited by 17
“This powerful machine learning technique has proven highly useful for providing solutions to a number of difficult computational problems ranging from vision to playing Go (Sejnowski, 2018, 2020; Silver et al. , 2018; Yang and Wang, 2020).”
MethodEvaluation of Human-AI Teams for Learned and Rule-Based Agents in HanabiPreprint
Siu, H. C., Peña, J., Chen, E., Zhou, Y., Lopez, V. J., Palko, K., Chang, K. C., & Allen, R. E. (2021) · arXiv (Cornell University) · doi:10.48550/arxiv.2107.07630 · Cited by 39
“Much of the contemporary work on learning-based agents has not evaluated such teaming metrics, due to their focus on single-player [3] or adversarial games [39, 43].”
BackgroundA Survey of Reasoning with Foundation ModelsPreprint
Sun, J., Zheng, C., Xie, E., Liu, Z., Chu, R., Liu, J., Xu, J., Ding, M., Li, H., Geng, M., et al. (2023) · OSF Preprints (OSF Preprints) · doi:10.31219/osf.io/ac4sp · Cited by 154
Systematic literature review on reinforcement learning in non-communicable disease interventions
Zhao, Y., Chaw, J. K., Liu, L., Chaw, S. H., Ang, M. C., & Tin, T. T. (2024) · Artificial Intelligence in Medicine · doi:10.1016/j.artmed.2024.102901 · Cited by 17
A survey on applications of reinforcement learning in spatial resource allocation
Zhang, D., Wang, M., Mango, J., Li, X., & Xu, X. (2024) · Computational Urban Science · doi:10.1007/s43762-024-00127-z · Cited by 19
“…learning (RL) methods have made breakthroughs in games (Kaiser et al., 2019; Lample & Chaplot, 2017; Littman, 1994), Go (Bouzy & Chaslot, 2006; Silver et al., 2018; Silver et al., 2007), autonomous driving (J. Chen et al., 2019a, 2019b; Kiran et al., 2021; Sallab et al., 2017; Shalev-Shwartz…”
Deep reinforcement learning-based resource allocation method for multi-satellite scheduling
Chen, X., Tian, T., Dai, G., Wang, M., Song, Z., & Xing, L. (2025) · Computers & Operations Research · doi:10.1016/j.cor.2025.107088 · Cited by 18
SINDy-RL for interpretable and efficient model-based reinforcement learning
Zolman, N., Lagemann, C., Fasel, U., Kutz, J. N., & Brunton, S. L. (2025) · Nature Communications · doi:10.1038/s41467-025-65738-4 · Cited by 71
Intuition, intelligence, data compression
Kipper, J. (2019) · Synthese · doi:10.1007/s11229-019-02118-8 · Cited by 18
Automation of Synthesized Optimal Control Problem Solution for Mobile Robot by Genetic Programming
Diveev, A., & Sofronova, E. (2019) · Advances in intelligent systems and computing · doi:10.1007/978-3-030-29513-4_77 · Cited by 15
Growiec, J. (2020) · Macroeconomic Dynamics · doi:10.1017/s1365100521000031 · Cited by 17
“…the recent decade we have witnessed a surge in artificial intelligence (AI) breakthroughs based on the methodology of deep neural networks [Tegmark (2017)], from autonomous vehicles and simultaneous language interpretation to self-taught superhuman performance at chess and Go [Silver et al. (2018)].”
Reinforcement learning applied to games
Crespo, J., & Wichert, A. (2020) · SN Applied Sciences · doi:10.1007/s42452-020-2560-3 · Cited by 17
“For instance, DeepMind used key aspects they learned while developing AlphaZero [36, 37], an agent capable of playing chess, go or shogi, to project AlphaFold’s [27] neural network architecture.”
BackgroundMethodInfluentialQuantum Continual Learning Overcoming Catastrophic Forgetting
Jiang, W., Lu, Z., & Deng, D.-L. (2022) · Chinese Physics Letters · doi:10.1088/0256-307x/39/5/050303 · Cited by 21
Strategies for Scaleable Communication and Coordination in Multi-Agent (UAV) Systems
Ponniah, J., & Dantsker, O. D. (2022) · Aerospace · doi:10.3390/aerospace9090488 · Cited by 17
Power System Fault Diagnosis Method Based on Deep Reinforcement Learning
Wang, Z., Zhang, Z., Zhang, X., Du, M., Zhang, H., & Liu, B. (2022) · Energies · doi:10.3390/en15207639 · Cited by 18
Polymer‐based neuromorphic devices: resistive switches and organic electrochemical transistors
Yamamoto, S. (2023) · Polymer International · doi:10.1002/pi.6520 · Cited by 15
Deep RL at Scale: Sorting Waste in Office Buildings with a Fleet of Mobile Manipulators
Herzog, A., Rao, K., Hausman, K., Lü, Y., Wohlhart, P., Yan, M., Lin, J., Arenas, M. G., Xiao, T. G., Kappler, D., et al. (2023) · Robotics: Science and Systems · doi:10.15607/rss.2023.xix.022 · Cited by 47
Krichen, M. (2023) · International Conference on Computing, Communication and Networking Technologies (ICCCNT) · doi:10.1109/icccnt56998.2023.10306453 · Cited by 35
“DNNs can automatically learn hierarchical representations of complex input data, s.a. images, sounds, and natural language, which makes them well-suited for handling high-dimensional state and action spaces [9].”
Toward More Human-Like AI Communication: A Review of Emergent Communication Research
Brandizzi, N. (2023) · IEEE Access · doi:10.1109/access.2023.3339656 · Cited by 43
“In the initial phase of AI research following the second AI winter, the focus was on identifying new areas where AI could outperform humans, with famous examples including chess [1], Go [2], and Starcraft [3].”
BackgroundInfluentialAdaptive instance similarity embedding for online continual learning
Han, Y.-N., & Liu, J. (2023) · Pattern Recognition · doi:10.1016/j.patcog.2023.110238 · Cited by 15
Deep reinforcement learning applied to statistical arbitrage investment strategy on cryptomarket
Vergara, G., & Kristjanpoller, W. (2024) · Applied Soft Computing · doi:10.1016/j.asoc.2024.111255 · Cited by 17
Trojan playground: a reinforcement learning framework for hardware Trojan insertion and detection
Sarihi, A., Patooghy, A., Jamieson, P., & Badawy, A. A. (2024) · The Journal of Supercomputing · doi:10.1007/s11227-024-05963-8 · Cited by 20
“RL agents have been extensively used in various application domains where decision making is required, e.g. , robotics control [41, 42], gaming [43, 44], autonomous driving [45], computer architecture [46, 47], and hardware security [48].”
Pitfalls (and advantages) of sophisticated large language models
Strasser, A. (2024) · Elsevier eBooks · doi:10.1016/b978-0-443-18851-0.00007-x · Cited by 20
Deep learning of first-order nonlinear hyperbolic conservation law solvers
Morand, V., Müller, N., Weightman, R., Piccoli, B., Keimer, A., & Bayen, A. M. (2024) · Journal of Computational Physics · doi:10.1016/j.jcp.2024.113114 · Cited by 22
Giacomoni, L., & Parisis, G. (2024) · IEEE Conference on Computer Communications (INFOCOM) · doi:10.1109/infocom52122.2024.10621288 · Cited by 26
“Reinforcement learning (RL) has been gaining substantial momentum in recent years, and increasingly sophisticated systems spanning a vast array of applications, such as games [1]– [4], natural language processing [5], and even nuclear fusion [6], have been developed.”
BackgroundLANTERN: Learning-Based Routing Policy for Reliable Energy-Harvesting IoT Networks
Taghizadeh, H., Safaei, B., Monazzah, A. M. H., Oustad, E., Lalani, S. R., & Ejlali, A. (2024) · IEEE Transactions on Network and Service Management · doi:10.1109/tnsm.2024.3450011 · Cited by 19
“MDP is a valuable alternative to existing heuristic techniques since it does not require extensive analytic modeling and can be used in highly complex dynamic systems [35], [36], [37], [38].”
Yan, Z., Li, L., Shang, J., & Zhao, H. (2024) · Neurocomputing · doi:10.1016/j.neucom.2024.128510 · Cited by 17
Multi-agent Reinforcement Learning for the Control of Three-Dimensional Rayleigh–Bénard Convection
Vasanth, J., Rabault, J., Alcántara-Ávila, F., Mortensen, M., & Vinuesa, R. (2024) · Flow Turbulence and Combustion · doi:10.1007/s10494-024-00619-2 · Cited by 34
“A successful application that contributed to its recent popularity is in gaming, where neural network-based agents are trained to play and win computer games such as Atari (Mnih et al. 2013), Go (Silver et al. 2018) and Starcraft (Vinyals et al. 2019).”
Engineering Topological States and Quantum‐Inspired Information Processing Using Classical Circuits
Chen, T., Zhang, W., Zou, D., Sun, Y., & Zhang, X. (2025) · Advanced Quantum Technologies · doi:10.1002/qute.202400448 · Cited by 21
Hippocampus shapes entorhinal cortical output through a direct feedback circuit
Butola, T., Hernández‐Frausto, M., Blankvoort, S., Flatset, M. S., Peng, L., Hairston, A., Johnson, C. D., Elmaleh, M., Amilcar, A., Hussain, F., et al. (2025) · Nature Neuroscience · doi:10.1038/s41593-025-01883-9 · Cited by 21
“4) led to our hypothesis that such reciprocal activity could iteratively refine cortical output and improve learning, akin to feedback loops enhancing performance in artificial neural networks in machine learning 40–42 .”
Deep Reinforcement Learning Based Power Control for Wireless Multicast Systems
Raghu, R., Upadhyaya, P. S., Panju, M., Agarwal, V., & Sharma, V. (2019) · Allerton Conference on Communication, Control, and Computing · doi:10.1109/allerton.2019.8919748 · Cited by 14
“Several, deep reinforcement learning techniques such as Deep Q-Network [13], Trust Region Policy Optimization (TRPO) [14], Proximal Policy Gradient (PPO) [15] etc. have been successfully applied to several large state-space dynamical systems such as Atari [16], AlphaGo [17] etc. DQN is one of the first DeepRL methods…”
MethodFinding Needles in a Moving Haystack: Prioritizing Alerts with Adversarial Reinforcement Learning
Tong, L., Lászka, Á., Yan, C., Zhang, N., & Vorobeychik, Y. (2020) · Proceedings of the AAAI Conference on Artificial Intelligence · doi:10.1609/aaai.v34i01.5442 · Cited by 24
“…for value and policy approximation has led to significant recent successes across several applications, most notably game-playing (Mnih et al. 2015; Silver et al. 2016; 2018), and a number of advances in recent years have made such approaches broadly applicable and highly effective (Van Hasselt,…”
BackgroundArtificial Intelligence, Chaos, Prediction and Understanding in Science
Sanjuán, M. A. F. (2021) · International Journal of Bifurcation and Chaos · doi:10.1142/s021812742150173x · Cited by 20
“Nature 529, 484-489 (2016) [5] Silver, D.”
BackgroundBanerji, S., Majumder, A., Hamrick, A. B., Menon, R., & Sensale‐Rodriguez, B. (2020) · Nano Communication Networks · doi:10.1016/j.nancom.2020.100312 · Cited by 17
Tackling Morpion Solitaire with AlphaZero-like Ranked Reward Reinforcement Learning
Wang, H., Preuß, M., Emmerich, M., & Plaat, A. (2020) · International Symposium on Symbolic and Numeric Algorithms for Scientific Computing (SYNASC) · doi:10.1109/synasc51798.2020.00033 · Cited by 13
“great success of the AlphaGo and AlphaZero programs [1], [3], [4] in two-player games, has inspired attempts in other”
BackgroundJao, M., Chan, S., Wu, M., & Lai, C. (2020) · The Journal of Physical Chemistry Letters · doi:10.1021/acs.jpclett.0c02393 · Cited by 15
Qi, N., Fan, Z., Huo, M., Du, D., & Zhao, C. (2020) · IEEE Transactions on Cybernetics · doi:10.1109/tcyb.2020.3040799 · Cited by 16
“In recent years, artificial intelligence has been successfully applied in complex decision-making fields, such as AlphaGo [30], [31]; AlphaZero [32], [33]; and AlphaStar [34].”
MethodDifferences Between Natural and Artificial Cognitive Systems
Singer, W. (2021) · Springer eBooks · doi:10.1007/978-3-030-54173-6_2 · Cited by 15
Solvent extraction process design using deep reinforcement learning
Plathottam, S. J., Richey, B., Curry, G., Cresko, J., & Iloeje, C. O. (2021) · Journal of Advanced Manufacturing and Processing · doi:10.1002/amp2.10079 · Cited by 15
Real-Time Decision Making for a Car Manufacturing Process Using Deep Reinforcement Learning
Gros, T. P., Groß, J., & Wolf, V. (2020) · Winter Simulation Conference · doi:10.1109/wsc48552.2020.9383884 · Cited by 19
“Deep reinforcement learning has been successfully applied to Atari games (Mnih et al. 2015; Mnih et al. 2013) and the games Go and Chess (Silver et al. 2016; Silver et al. 2017; Silver et al. 2018).”
MethodEstimation Error Correction in Deep Reinforcement Learning for Deterministic Actor-Critic Methods
Sağlam, B., Duran, E., Cicek, D. C., Mutlu, F. B., & Kozat, S. S. (2021) · IEEE International Conference on Tools with Artificial Intelligence (ICTAI) · doi:10.1109/ictai52525.2021.00027 · Cited by 17
Generalization in Mean Field Games by Learning Master Policies
Perrin, S., Laurière, M., Pérolat, J., Élie, R., Geist, M., & Pietquin, O. (2022) · Proceedings of the AAAI Conference on Artificial Intelligence · doi:10.1609/aaai.v36i9.21173 · Cited by 53
“…1959, Samuel, 1959], most of recent breakthroughs remain limited to a small number of players, e.g. , for chess [Campbell et al., 2002], Go [Silver et al., 2016, 2017, 2018], poker [Brown and Sandholm, 2018, Moravˇc´ık et al., 2017] or even video games such as Star-craft [Vinyals et al.,…”
Song, S., Chen, H., Sun, H., Liu, M., & Xia, T. (2022) · IEEE Robotics and Automation Letters · doi:10.1109/lra.2021.3139950 · Cited by 15
“Asamodel-basedreinforcementlearning(RL)method,which beat a human champion in the game of Go, MCTS [18]–[20] requires a problem-specific reactive experience and long-term value estimation.”
BackgroundInterleaving Monte Carlo Tree Search and Self-Supervised Learning for Object Retrieval in Clutter
Huang, B., Guo, T., Boularias, A., & Yu, J. (2022) · Proceedings - IEEE International Conference on Robotics and Automation/Proceedings · doi:10.1109/icra46639.2022.9812132 · Cited by 24
“But, perhaps the most prominent line of work in reinforcement learning [3] that closely aligns with this paradigm is the application of Monte Carlo Tree Search (MCTS) for carrying out self-supervised learning in games [4], [5], where an “understanding” of a game emerges from a lifelong self-play and is gradually disti…”
BackgroundRaheb, M. A., Niazmand, V. R., Eqra, N., & Vatankhah, R. (2022) · Computers in Biology and Medicine · doi:10.1016/j.compbiomed.2022.105860 · Cited by 18
Graph Neural Networks for Voltage Stability Margins With Topology Flexibilities
Guddanti, K. P., Weng, Y., Marot, A., et al. (2022) · IEEE Open Access Journal of Power and Energy · doi:10.1109/oajpe.2022.3223962 · Cited by 15
“It also helps in estimating the power grid margin (D-VSI) instantaneously which helps in CPU intensive applications such asMonte Carlo Tree Searchbased (MCTS) solutions (like AlphaGo) for power grids with margin constraints [51], [52] (these approaches are currently impractical for power grids due to the dependency on…”
BackgroundIntelligent Buses in a Loop Service: Emergence of No-Boarding and Holding Strategies
Saw, V.-L., Vismara, L., & Chew, L. Y. (2020) · Complexity · doi:10.1155/2020/7274254 · Cited by 16
Lee, H.-S., & Lee, J.-W. (2020) · IEEE Access · doi:10.1109/access.2020.3037323 · Cited by 15
“In particular, deep reinforcement learning (RL) has been widely used for various complex problems such as robot control [26], playing the game of Go [27],”
MethodFrom self-tuning regulators to reinforcement learning and back againPreprint
Matni, N., Proutière, A., Rantzer, A., et al. (2019) · arXiv (Cornell University) · doi:10.48550/arxiv.1906.11392 · Cited by 15
“ol. I. INTRODUCTION & MOTIVATION With their recent successes in image classification, video game playing [1], sophisticated robotic simulations [2], [3], and complex strategy games such as Go [4], [5], machine and reinforcement learning (RL) are now being applied to plan and control the behavior of autonomous syste…”
BackgroundMethodKnowledge Induced Deep Q-Network for a Slide-to-Wall Object Grasping.Preprint
Liang, H., Lou, X., & Choi, C. (2019) · arXiv (Cornell University) · Cited by 14
Algorithm, Human, or the Centaur: How to Enhance Clinical Care?Preprint
Dean, A., Orfanoudaki, A., Saghafian, S., Song, K., Chakkera, H. A., & Cook, C. B. (2022) · SSRN Electronic Journal · doi:10.2139/ssrn.4302002 · Cited by 38
“In the field of reinforcement learning, several algorithms have achieved superior performance compared to human experts, defeating the world’s best players in cerebral games (see, e.g., (Silver et al. 2018)).”
BackgroundAutomated classification of cellular expression in multiplexed imaging data with Nimbus
(2025) · doi:10.1038/s41592-025-02826-9 · Cited by 24
Convolutional Neural Networks for Raw Signal Classification in CNC Turning Process Monitoring
Stathatos, E., Tzimas, E., Benardos, P., & Vosniakos, G.-C. (2024) · Sensors · doi:10.3390/s24051390 · Cited by 19
Machine Memory Intelligence: Inspired by Human Memory Mechanisms
Zheng, Q., Liu, H., Zhang, X., Yan, C., Cao, X., Gong, T., Liu, Y., Shi, B., Peng, Z., Fan, X., et al. (2025) · Engineering · doi:10.1016/j.eng.2025.01.012 · Cited by 17
Deep learning in next-generation vaccine development for infectious diseases
Bhattacharya, M., Lo, Y.-H., Chatterjee, S., Das, A., Wen, Z., & Chakraborty, C. (2025) · Molecular Therapy — Nucleic Acids · doi:10.1016/j.omtn.2025.102586 · Cited by 21
A Comprehensive Survey of Data Augmentation in Visual Reinforcement Learning
Ma, G., Ding, Z., Yuan, Z., Wang, X., Yuan, B., & Tao, D. (2025) · International Journal of Computer Vision · doi:10.1007/s11263-025-02472-w · Cited by 71
“This paradigm has been successfully employed in video games [151], autonomous driving [82], robot control [76], and other areas.”
Learning Task-Distribution Reward Shaping with Meta-Learning
Zou, H., Ren, T., Yan, D., Su, H., & Zhu, J. (2021) · Proceedings of the AAAI Conference on Artificial Intelligence · doi:10.1609/aaai.v35i12.17337 · Cited by 17
“Credit assignment (Minsky 1961) remains a key challenge on the learning efficiency (sample complexity) of Reinforcement Learning (RL) (Mnih et al. 2015; Silver et al. 2018), which usually requires many samples from the environment to learn well.”
BackgroundRecent advances in leveraging human guidance for sequential decision-making tasks
Zhang, R., Torabi, F., Warnell, G., & Stone, P. (2021) · Autonomous Agents and Multi-Agent Systems · doi:10.1007/s10458-021-09514-w · Cited by 31
Modified action decoder using Bayesian reasoning for multi-agent deep reinforcement learning
Du, W., Ding, S., Zhang, C., & Du, S. (2021) · International Journal of Machine Learning and Cybernetics · doi:10.1007/s13042-021-01385-7 · Cited by 14
Free-Space Optical Neural Network Based on Optical Nonlinearity and Pooling Operations
Sadeghzadeh, H., Koohi, S., & Paranj, A. F. (2021) · IEEE Access · doi:10.1109/access.2021.3123230 · Cited by 18
“INTRODUCTION In today’s world, machine learning technology, specifically deep neural network, plays a key role in various applications [1] such as natural language processing [2], medicine [3], and game playing [4] to name but a few.”
BackgroundSharma, S., Rawal, Y. S., Pal, S., & Dani, R. (2022) · Lecture notes in networks and systems · doi:10.1007/978-981-16-5655-2_48 · Cited by 17
Underwater Target Detection Based on Reinforcement Learning and Ant Colony Optimization
Wang, X., Zhu, Y., Li, D., et al. (2022) · Journal of Ocean University of China · doi:10.1007/s11802-022-4887-4 · Cited by 14
“The problem to be solved is how to choose the action that can achieve the optimal goal through learning (Silver et al ., 2018).”
BackgroundRisk-sensitive policies for portfolio management
Wang, M., & Ku, H. (2022) · Expert Systems with Applications · doi:10.1016/j.eswa.2022.116807 · Cited by 18
Reinforcement learning: A brief guide for philosophers of mind
Haas, J. (2022) · Philosophy Compass · doi:10.1111/phc3.12865 · Cited by 23
“…system was able to develop superhuman play in the game Go by combining supervised learning of human games of Go and reinforcement learning that was used to improve the foregoing policy by having the system play itself (Silver et al., 2016; see also Silver et al., 2017; Silver et al., 2018).”
MethodNeuroevolution-Based Generation of Tests and Oracles for Games
Feldmeier, P., & Fraser, G. (2022) · IEEE/ACM International Conference on Automated Software Engineering (ASE) · doi:10.1145/3551349.3556939 · Cited by 16
“Many reinforcement learning techniques have been shown to be capable of learning to play different game genres such as strategy games [55, 56], arcade games [63], first-person shooter games [37, 42], and casual games [33].”
BackgroundUnveiling the Relationship between Economic Growth and Equality for Developing Countries
Cheong, T. S., Wan, G., & Chui, D. K. H. (2022) · China & World Economy · doi:10.1111/cwe.12435 · Cited by 15
Fan, Z., Huo, M., Quarta, A. A., Mengali, G., & Qi, N. (2022) · Aerospace Science and Technology · doi:10.1016/j.ast.2022.107946 · Cited by 14
Reinforcement Learning Toolkits for Gaming: A Comparative Qualitative Analysis
Mekni, M., Jayaramireddy, C. S., & Naraharisetti, S. V. V. S. S. (2022) · Journal of Software Engineering and Applications · doi:10.4236/jsea.2022.1512024 · Cited by 14
Song, Z., Zhou, Q., Lu, S., Dieb, S., Ling, C., & Wang, J. (2023) · The Journal of Physical Chemistry Letters · doi:10.1021/acs.jpclett.3c00242 · Cited by 15
Human achievement and artificial intelligence
Karlan, B. (2023) · Ethics and Information Technology · doi:10.1007/s10676-023-09713-x · Cited by 15
“2 More recent work has attempted to replicate superhuman performance at multiple games without hand-coding expertise of any kind (even the rules of the game) into the algorithm (Silver et al., 2018).”
Background
Metadata
- Authors
- Silver, D., Hubert, T., Schrittwieser, J., Antonoglou, I., Lai, M., Guez, A., Lanctot, M., Sifre, L., Kumaran, D., Graepel, T., Lillicrap, T., Simonyan, K., & Hassabis, D.
- Affiliations
- Google DeepMind (United Kingdom) (GB)
- Corresponding author
- Silver, David
- Hubert, Thomas
- Schrittwieser, Julian
- Hassabis, Demis
- ORCID
- Venue
- Science
- Published
- 2018-12-07
- Volume
- 362
- Issue
- 6419
- Pages
- 1140-1144
- ISSN
- 0036-8075 (ISSN-L)
- 1095-9203
- OpenAlex
- W2902907165
- Cited by
- 3,562
- References
- 30
- Citations by year
- 2026: 257 · 2025: 427 · 2024: 512 · 2023: 514 · 2022: 569 · 2021: 546 · 2020: 462 · 2019: 256 · 2018: 13
- FWCI
- 179.89
- Type
- journal article
- Language
- en
- Open access
- bronze
- doi.org
- Locations
- Science (publishedVersion)
- PubMed (publishedVersion)
- UCL Discovery (University College London) (submittedVersion)
- Funders
- DeepMind
- Topics
- Physical Sciences › Computer Science › Artificial Intelligence › Artificial Intelligence in Games
- Physical Sciences › Computer Science › Artificial Intelligence › Reinforcement Learning in Robotics
- Physical Sciences › Computer Science › Artificial Intelligence › Artificial Intelligence Applications
- Keywords
- AlphaZero, chess, shogi, Go, evaluation function learning, game tree search
- MeSH terms
- Algorithms, Artificial Intelligence, Humans, Reinforcement, Psychology, Software, Video Games
About these checks
Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.
The claim verdicts this report counts are experimental: their accuracy is not yet measured.
A language model drew the claims from the paper on 2026-10-10, under version 3 of the claim definition (model deepseek-v4.1-flash, prompt extract-v3).
A second model labelled each claim on 2026-10-10 (model jev-1.13-20260917, definition revision gold-guideline-v0.2).
Cited sources are checked with a language model’s help (gpt-6-sol).
Statistics and arithmetic are recomputed, not judged.
Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.
What counts as a claim ›How we measure confidence ›How we check this ›
Are you an author, or do you see a problem with this page? Request a review
Data sources
Bibliographic data: Semantic Scholar, OpenAlex, Crossref.
Contains information from the Semantic Scholar Academic Graph Datasets, made available under the ODC Attribution License (ODC-BY 1.0).
Semantic Scholar’s data is described in: Kinney, R., et al. (2023). The Semantic Scholar Open Data Platform. arXiv:2301.10140.