要旨
Date and Time: March 1st 3:00pm-5:00pm (JST)
Venue: Zoom webinar
Language: English
3:00pm – 4:00pm
Speaker: Aurelio Cortese (Advanced Telecommunication Research Institute International, ATR)
Title: Metacognitive resources for efficient learning
Abstract: A central issue in reinforcement learning (RL) is the ‘curse-of-dimensionality’, arising when the degrees-of-freedom of the problem are much larger than the number of available training samples. In the brain, higher cognitive functions such as metacognition or abstraction can provide a biological solution by generating low dimensional representations on which RL can operate. In this talk I will discuss our work in which we used neuroimaging and computational modeling to investigate the neural basis of efficient RL. We found that people can learn remarkably complex task structures non-consciously, but also that metacognition appears tightly coupled to this learning ability. Furthermore, when people use an explicit (conscius) policy to select relevant information, learning is accelerated by abstractions. The prefrontal part of the brain is differentially involved in separate aspects of learning: dorsolateral prefrontal cortex pairs with metacognitive processes, while ventromedial prefrontal cortex with valuation and abstraction. I will discuss the implications of these findings and focus on a new model framework toward metacognitive AI.
Reference:
From internal models towards metacognitive AI, (Biol. Cyb., 2021) M. Kawato, A. Cortese [paper]
Unconscious reinforcement learning of hidden brain states supported by confidence, (Nature Comm., 2020) A. Cortese, H. Lau, M. Kawato [paper]
Value signals guide abstraction in learning, (eLife, 2021) A. Cortese, A. Yamamoto, M. Hashemzadeh, P. Sepulveda, M. Kawato, B. De Martino [paper]
Bio: Aurelio Cortese is the deputy head of Dept. of Decoded Neurofeedback at Advanced Telecommunications Research Institute International (ATR) in Kyoto. He is also a principal investigator of the ERATO-Ikegaya project and a visiting scientist at the Institute of Cognitive Neuroscience, University College London (UK). He finished his master course study at Ecole Polytechnique Fédérale de Lausanne (EPFL), and got his PhD in computational neuroscience from Nara Institute of Science and Technology (NAIST), Japan in 2016.
The long-term goal of Aurelio’s research is to understand the nature of flexible and adaptive behaviors in humans and biological agents in general. In particular, he has been interested in higher-order functions in reinforcement learning, metacognition and consciousness.
4:00pm – 5:00pm
Speaker: Emtiyaz Khan (RIKEN AIP)
Title: The Bayesian Learning Rule
Abstract: Humans and animals have a natural ability to autonomously learn and quickly adapt to their surroundings. How can we design AI systems that do the same? In this talk, I will present Bayesian principles to bridge such gaps between humans and AI. I will show that a wide-variety of machine-learning algorithms are instances of a single learning-rule called the Bayesian learning rule. The rule unravels a dual perspective yielding new mechanisms for knowledge transfer in machine-learning based AI systems. My hope is to convince the audience that Bayesian principles are indispensable for an AI that learns as efficiently as we do.
Reference: The Bayesian Learning Rule, (Preprint) M.E. Khan, H. Rue [ arXiv ] [ Tweet ]
Bio: Emtiyaz Khan (also known as Emti) is a team leader at the RIKEN center for Advanced Intelligence Project (AIP) in Tokyo where he leads the Approximate Bayesian Inference Team. He is also an external professor at the Okinawa Institute of Science and Technology (OIST). Previously, he was a postdoc and then a scientist at Ecole Polytechnique Fédérale de Lausanne (EPFL), where he also taught two large machine learning courses and received a teaching award. He finished his PhD in machine learning from University of British Columbia in 2012. The main goal of Emti’s research is to understand the principles of learning from data and use them to develop algorithms that can learn like living beings. For more than a decade, his work has focused on developing Bayesian methods that could lead to such fundamental principles. The approximate Bayesian inference team now continues to use these principles, as well as derive new ones, to solve real-world problems.
All participants are required to agree with the AIP Seminar Series Code of Conduct.
Please see the URL below.
https://aip.riken.jp/event-list/termsofparticipation/?lang=en
RIKEN AIP will expect adherence to this code throughout the event. We expect cooperation from all participants to help ensure a safe environment for everybody.
詳細情報
日時 | 2022/03/01(火) 15:00 - 17:00 |
URL | https://c5dc59ed978213830355fc8978.doorkeeper.jp/events/133661 |