Our website uses cookies. By continuing to use this site, you agree to our use of cookies in accordance with this notice and the Terms of Use.
JOURNAL EXPERTS OF THE JOURNAL CONFERENCES TELEGRAM FOR AUTHORS
Submit manuscript
Home / Conferences / VII All-Russian scientific and practical conference with international participation "Sports Informatics Day" / Collection of abstracts of the VII All-Russian Conference with international participation scientific and practical conference "SPORTS INFORMATICS DAY" on December 4-5, 2023, Moscow
APPROBATION OF DETECTED WEAKNESS OF SELF-TAUGHT GO GAME AI KATAGO
Submit manuscript PDF Text
Indexing and databases
Istina MSU → RINC
UDC
00 7 004 796 796/799
CSCSTI
28.23 28.17 28.01 77.05 77.29
7 more Collapse
77.31 80.37 20.53 20.01 50.33 50.41 30.51
Russian Classification of Professions by Education
02.04.01 02.06.01 31.08.39 49.02.01 49.06.01
1 more Collapse
12.02.06
Russian Trade and Bibliographic Classification
2352 4614 4618 4612 2301
7 more Collapse
3015 5106 518 5182 5185 762 7623
Russian Library and Bibliographic Classification
75 6539 73 606
BISAC
COM059000 COM012000 COM014000 COM072000 COM017000
49 more Collapse
COM062000 COM018000 COM021030 COM071000 COM023000 COM025000 COM074000 COM032000 COM077000 BUS021000 GAM001000 GAM016000 GAM004050 EDU039000 EDU036000 EDU037000 EDU027000 MED084000 MED078000 MED076000 HEA046000 HEA006000 HEA034000 HEA002000 HEA017000 HEA007000 HEA010000 HEA038000 MAT011000 MAT029030 MAT025000 SPO003000 SPO068000 SPO077000 SPO061020 SPO061030 SPO011000 SPO063000 SPO073000 SPO015000 SPO016000 SPO075000 SPO024000 SPO040000 SPO066000 SPO047000 SPO051000 PSY012000 PSY042000
VAK Russia
5.8.4 5.8.5 5.8.6 5.8.7 5.2.2
10 more Collapse
1.1.10 1.2.1 1.2.2 1.2.3 1.5.5 2.2.11 2.2.12 2.3.1 2.3.4 2.3.8
APPROBATION OF DETECTED WEAKNESS OF SELF-TAUGHT GO GAME AI KATAGO
Type:
Conference abstract
DOI:
https://doi.org/10.62105/2949-6349-2024-1-S1-97-99
EDN:
https://elibrary.ru/igicho
Pages:
97–99
Published:
08.02.2024
Language:
Russian, English
Abstract:
The article describes testing a weakness of the go-playing AI KataGo (9d) detected by K. Pelrine et al. Building groups following the developed 4-level algorithm reveals the incapability of AI to understand intentional sacrifice of larger groups and to prioritize when discovering several similar sacrifices on the board. The value of the developed strategy lies in human victory over a potentially unbeatable program even regarding the fact that human is playing on the level of 1q (2000 Elo).

Keywords:
go game, baduk, weiqi, KataGo, neural networks
Text (PDF): Read Download

1. AlphaGo by Google DeepMind. URL: https://www.deepmind.com/research/highlighted-research/alphago (accessed 20.09.2023).

2. KataGo on GitHub. URL: https://github.com/lightvector/KataGo/releases?q=v1.12.4&expanded=true (accessed 20.09.2023).

3. Leela official website. URL: https://www.sjeng.org/leela.html (accessed 20.09.2023).

4. Lauder E. Tencent's Made a Go-Playing AI to Rival Google's AlphaGo. URL: https://aibusiness.com/companies/tencent-s-made-a-go-playing-ai-to-rival-google-s-alphago (accessed 20.09.2023).

5. Wang T.T., Gleave A., Tseng T., Pelrine K., Belrose N. et al. Adversarial policies beat superhuman Go AIs // arXiv:2211.00241 [cs.LG]. URL: https://goattack.far.ai/adversarial-policy-katago (accessed 20.09.2023). DOI:https://doi.org/10.48550/arXiv.2211.00241

6. Financial Times. Man beats machine at Go in human victory over AI. URL: https://www.ft.com/content/175e5314-a7f7-4741-a786-273219f433a1 (accessed 27.11.2023).

© rjits.ru
Agreement Personal data protection and processing policy Support Powered by Editorum, 2026 Instructions
Login or Create
* Forgot your password?