June 26, 2017 - CogSci London (Windsor Room)

How do we understand, act with and learn from others and how might we design intelligent agents that can do the same? This full-day workshop will feature new approaches to studying social intelligence in humans and machines from cognitive, developmental and engineering perspectives. The workshop will focus on theories and behavioral experiments that shed light on (1) multi-agent cooperation and coordination, (2) mental state inferences and prediction, and (3) social and cultural learning through observation, teaching, and communication.

Schedule

Speaker

Talk Title


Introductory Remarks

Nick Chater

University of Warwick

The Social Contract in Miniature: The virtual bargaining theory of social interaction

Natalie Sebanz

Central European University

How t(w)o act together: Coordination mechanisms in joint action


Igor Mordatch

OpenAI & CMU

Lunch

Emergence of Grounded Compositional Language in Multi-Agent Populations

Stuart Russell

UC Berkeley

Cooperative inverse reinforcement learning


Hyo Gweon

Stanford University

Coffee Break

Learning to learn from others: Inference, evaluation, and communication in social learning


Joel Leibo

Google DeepMind

A multi-agent reinforcement learning model of common-pool resource appropriation

Victoria Southgate

University of London Birkbeck

Early social intelligence: how do infants understand others?


Speakers w/

Elizabeth Spelke

Harvard University

Panel discussion