検索条件

キーワード
タグ
ツール
開催日
こだわり条件

タグ一覧

JavaScript
PHP
Java
Ruby
Python
Perl
Scala
Haskell
C言語
C言語系
Google言語
デスクトップアプリ
スマートフォンアプリ
プログラミング言語
U/UX
MySQL
RDB
NoSQL
全文検索エンジン
全文検索
Hadoop
Apache Spark
BigQuery
サーバ構成管理
開発サポートツール
テストツール
開発手法
BI
Deep Learning
自然言語処理
BaaS
PaaS
Iaas
Saas
クラウド
AI
Payment
クラウドソフトウェア
仮想化ソフトウェア
OS
サーバ監視
ネットワーク
WEBサーバ
開発ツール
テキストエディタ
CSS
HTML
WEB知識
CMS
WEBマーケティング
グラフィック
グラフィックツール
Drone
AR
マーケット知識
セキュリティ
Shell
IoT
テスト
Block chain
知識

[Team ABI Seminar] Lucy Lai, Policy compression: acting with limited cognitive resources

2025/06/06(金)
02:00〜04:00

主催:RIKEN AIP Public

This talk will be held in a hybrid format, both in person at AIP Open Space of RIKEN AIP (Nihonbashi office) and online by Zoom. AIP Open Space: *only available to AIP researchers.

DATE, TIME & LOCATION
Friday, June 6th, 11:00 - 13:00, RIKEN AIP Nihombashi Office, Open Space

TITLE
Policy compression: acting with limited cognitive resources

ABSTRACT
Decision making taxes cognitive resources. For example, when deciding where to eat for lunch, we must consider what food options are available to us at each restaurant. But time constraints or mental fatigue can often steer us towards habitual choices, such as eating at the same ramen shop every day. As illustrated here, our behavior is undeniably shaped by constraints, or limits, on our cognitive resources. To better understand how this resource limitation affects behavior, I use tools from reinforcement learning and information theory to build testable models of action selection, explaining behavioral phenomena such as choice biases and the use of strategies like "chunking" to reduce mental load. In the second half of the talk, I will highlight a potential neural signature of cost-sensitive action selection in the basal ganglia. Lesion studies in rats reveal that distinct regions of the striatum collaborate to balance cognitively demanding decisions with more automatic, habitual actions. By integrating insights from theory, behavior, and neuroscience, I hope to paint a vivid picture of how our decisions and actions are shaped by the brain’s drive to balance reward and cognitive costs.

BIO
Lucy Lai is a TSVP Visiting Scholar at the Okinawa Institute of Science and Technology (OIST) in Okinawa, Japan, and an incoming Assistant Teaching Professor at the University of California, San Diego (UCSD). Her research interests span learning and decision making, belief change, and human-AI collaboration. Lucy recently completed her Ph.D. in Neuroscience at Harvard University, where she was advised by Sam Gershman. Beyond research, Lucy enjoys teaching and mentoring, writing poetry, and latin dancing.

Workship