Research talk: Maia Chess: A human-like neural network chess engine
2 234
9.3
Microsoft Research335 тыс
Следующее
18.02.22 – 5 8585:37Популярные
Опубликовано 8 февраля 2022, 18:07
Speaker: Reid McIlroy-Young, PhD Student, University of Toronto
Even when machine learning surpasses human ability in a domain, there are many reasons why AI systems that capture human-like behavior would be desirable. For example, humans may want to learn and collaborate, or humans may need to interact with independent AI agents. A first step in aligning AI agents’ behavior to that of humans is creating agents that better understand human behavior. University of Toronto PhD student Reid McIlroy-Young will present his work, done in collaboration with Microsoft Research, building neural chess engines that can predict human behavior at different skill levels. Furthermore, these engines can be calibrated to target the decisions of specific players via fine-tuning. He will first discuss the value of studying the intersection of human and AI, and the results. He will also discuss where reinforcement learning was outperformed by a classification approach and conclude with a look at the benefits of having research that can be understood by a wide audience.
Learn more about the 2021 Microsoft Research Summit: Aka.ms/researchsummit
Even when machine learning surpasses human ability in a domain, there are many reasons why AI systems that capture human-like behavior would be desirable. For example, humans may want to learn and collaborate, or humans may need to interact with independent AI agents. A first step in aligning AI agents’ behavior to that of humans is creating agents that better understand human behavior. University of Toronto PhD student Reid McIlroy-Young will present his work, done in collaboration with Microsoft Research, building neural chess engines that can predict human behavior at different skill levels. Furthermore, these engines can be calibrated to target the decisions of specific players via fine-tuning. He will first discuss the value of studying the intersection of human and AI, and the results. He will also discuss where reinforcement learning was outperformed by a classification approach and conclude with a look at the benefits of having research that can be understood by a wide audience.
Learn more about the 2021 Microsoft Research Summit: Aka.ms/researchsummit
Свежие видео
8 дней – 6017:22
9 дней – 301 24228:45
9 дней – 3310:26
10 дней – 3452:08
10 дней – 21 0460:30
15 дней – 3580:28Случайные видео
261 день – 2 6145:27
351 день – 15 63930:51
31.07.23 – 189 7982:28
18.05.23 – 157 3950:35
15.02.22 – 106 8677:11
14.09.20 – 26 8711:06
10 часов – 1370:20
3 дня – 141 78614:34
6 дней – 67210:59
7 дней – 834:07
8 дней – 3011:25
8 дней – 4382:45
10 дней – 12 0090:31
2 дня – 1611:03
