Faculty

Tom Griffiths

Tom Griffiths, Lab Director

(webpage)


Postdocs

Robert Hawkins

Robert Hawkins

(webpage) I am interested in the cognitive mechanisms that allow people to flexibly coordinate and communicate with one another. My work uses computational models and multi-agent experiments to test theories of learning and adaptation in social interaction.


Mark Ho

Mark Ho

(webpage) Broadly, I am interested in how minds interface with other minds to produce complex social behaviors. Humans have a highly developed capacity to reason about and represent what others want and how they think, and this is central to our forms of communication, competition, coordination, and culture. My work has explored this capacity in non-verbal domains, such as teaching with reward/punishment, teaching by intervention, and teaching by demonstration, with the aim of developing a computational account of folk pedagogy. Currently, I am exploring how compositional and hierarchical representations influence and are influenced by a context that includes other intentional agents. In investigating these phenomena, I draw on several formalisms including game theory, reinforcement learning, and Bayesian probability.


Thomas Langlois

Thomas Langlois

(webpage) Vision is an active process. Far from being passive recipients of external information, our visual systems are constantly generating meaning by combining sensory information with internal beliefs about the structure of the world around us. From the perspective of Bayesian statistics, these beliefs correspond to perceptual priors. My research interests center around uncovering the structure of these priors and understanding the role that they play in perception and cognition. For a representative example of my current work, see our most recent publication in PNAS. I use innovative experimental methods combined with Bayesian computational modeling in my work. I completed my Ph.D. in Thomas Griffiths’ Computational Cognitive Science Lab at UC Berkeley in August of 2018. Prior to completing my Ph.D., I completed an M.S. in Computer Science (EECS), also at UC Berkeley. I am currently at Princeton University working as a postdoctoral researcher in the Computational Cognitive Science Lab in the Computer Science Department. I am also affiliated with the UC Berkeley Department of Psychology and the Computational Auditory Perception Research Group in the Max Planck Institute for Empirical Aesthetics.


Joshua Peterson

Joshua Peterson

My research employs machine learning and large datasets as tools to understand human cognition and predict behavior. In many domains, machine learning models are the only models that approach human performance for complex naturalistic tasks, and thus provide candidate models of cognition that can be critiqued and improved. My most recent projects explore how machine learning can be used to supplement the ingenuity of researchers by automating the search for interpretable theories of human behavior.


Ilia Sucholutsky

Ilia Sucholutsky

(webpage) I’m fascinated by deep learning and its ability to reach superhuman performance on so many different tasks. I want to better understand how neural networks achieve such impressive results… and why sometimes they don’t. Recently, I've been focused on improving deep learning in small data settings. The current paradigm in AI research is to train large models on large datasets using massive computational resources. While this trend does lead to improvements in predictive power, it leaves behind the multitude of researchers, companies, and practitioners who do not have access to sufficient funding, compute power, or volume of data. I'm interested in developing data-efficient methods that can help rectify this growing divide.


Bas van Opheusden

Bas van Opheusden

I'm interested in how people make decisions and learn strategies for complex decision-making tasks such as board games. I'm fascinated by how these kind of games can have simple rules but still require sophisticated strategies to play well or optimally, and how people learn such strategies from experience or observation. I'm currently working on a project exploring observation learning and cumulative evolution in a sorting game (with Bill Thompson), and a number of projects related to testing resource rationality in information-gathering tasks (with Fred Callaway).


Graduate Students

Mayank Agrawal

Mayank Agrawal

(webpage) Broadly, I want to understand the factors that enable the flexibility of human cognition. In one line of research, I leverage machine learning algorithms to posit representations that humans could use to solve various cognitive tasks. In another, I use the language of optimization theory to elucidate the control mechanisms that underlie memory, learning, and decision-making. My research draws heavily from computer science, psychology, neuroscience, and philosophy.


Ruairidh McLennan Battleday

Ruairidh McLennan Battleday

(webpage) In my research, I study generalization: how our inference about the novel and unknown is guided by our evolved and encountered past. This entails studying and formalizing generalization and analogical learning in humans, and testing these ideas by using them to create better machine-learning algorithms. More broadly, I am interested in furthering our understanding of cognition and intelligence by uniting insights from high-level theories and ideologies of the brain, mind, and computation.


Fred Callaway

Fred Callaway

(webpage) Intelligent agents must continually respond to and learn from their environment. Mathematical models from Bayesian statistics and reinforcement learning can provide optimal solutions to these problems; but they are often intractable to compute. How do humans find good approximations to these optimal solutions using limited neural resources? In particular, how do they balance the competing goals of learning, deciding, and conserving resources? I aim to study this question with game-based empirical experiments and computational models inspired by machine learning algorithms.


Michael Chang

Michael Chang

(webpage) I am interested in understanding and building machines that generalize as humans do. I have recently been exploring problems related to learners with compositional representations and computations. The goal is to work towards building agents that design and update their own languages for modeling the world and solving problems. I hypothesize that such learners would need to 1) decompose their percepts into a set of coherent representations that can be internally and separately manipulated (e.g. discover and represent objects in a scene), and 2) encapsulate computations that can be reused across various scenarios, including ones not encountered by the agent. Areas of research that I draw from include deep learning, reinforcement learning, and program induction.


Rachit Dubey

Rachit Dubey

(webpage) Curiosity is one of the hallmarks of human intelligence and is crucial to scientific discovery and invention - yet our understanding of curiosity remains quite limited. What is the function of curiosity? How does curiosity develop? My goal is to better understand such aspects of curiosity and also explore how curiosity relates to cognitive processes such as creativity and metareasoning. By studying curiosity through a computational lens, I intend to develop a better theoretical foundation of curiosity which can then lead to applications in various pedagogical settings.


Erin Grant

Erin Grant

(website) I connect methodology from cognitive science with technology from machine learning in order to identify and ameliorate cases in which our current understanding of human and machine intelligence—and their interactions—falls short. I focus on situations in which humans and machine learning systems need to generalize on the basis of insufficient evidence, and aim to investigate, compare, and intervene on the inductive priors that enable these systems to do so. I explore a variety of domains through this lens, including language learning, category learning, and task representation, using a combination of computational modeling and behavioral experiments.


Matt Hardy

Matt Hardy

(website) Every day people encounter a neverending set of complicated decisions, difficult tradeoffs, and unforeseeable developments. How do people navigate this complexity and uncertainty? I study this question using computational modeling, behavioral experiments, and observational data analysis. I am especially interested in investigating psychological phenomena in individuals situated in social networks and groups. Cognition is often studied as an isolated process and uses results from simple, individual-level experiments to predict behavior in real-world domains. However, people rarely make decisions in isolation, and many of life’s dilemmas would be impossible or intractable to solve alone. A better understanding of the relationship between individual and group cognition is key to understanding how people thrive in the complexity and uncertainty the real world presents.


Sreejan Kumar

Sreejan Kumar

(website) One hallmark of human cognition is the ability to form abstract representations to solve complex problems with relatively small amounts of data and strongly generalize these abstractions to other problems. Cognitive scientists have said that the acquisition of these abstractions can be modeled by Bayesian inference over discrete, symbolic, and structured representations such as graphs, grammars, predicate logic, programs, etc. However, some cognitive scientists have argued that abstract knowledge can be modeled as emergent phenomenon from statistical learning of distributed, sub-symbolic systems with relatively unstructured representations. Modern deep learning research have developed a variety of architectures for distributed, sub-symbolic systems that can solve difficult tasks, but the conditions in which these systems can emerge human-like abstractions is unclear. I am interested in combining symbolic probabilistic models with novel cognitive tasks that require abstract problem solving to formally characterize how humans acquire and utilize abstract knowledge. I am also interested in utilizing the same paradigm to figure out the architectures and training regimes in which modern deep learning systems can solve these problems with human-like abstractions.


Raja Marjieh

Raja Marjieh

How do humans derive semantic representations from perceptual objects? What are the computational principles underlying their structure? How can we characterize them? In my research, I engage with these problems by leveraging large-scale online experiments and designing paradigms that implement a human instantiation of various algorithms from physics and machine learning. I am also interested in understanding how these representations are modulated by social interactions, especially in the context of creative and aesthetic processes, such as what constitutes a pleasant chord or melody.


Kerem Oktar

Kerem Oktar

(website) My research aims to clarify the psychological and computational basis of disagreement – across scales, domains, and agents – from definition to intervention. I also study decision-making; in particular, I am interested in understanding people's preferences for relying on intuition vs. deliberation. I take a two-pronged approach to studying these questions. To generate theories, I combine insights from analytical philosophy, probability theory, and empirical psychology. To test these theories, I use behavioral experiments, computational models, and statistics.


Ted Sumers

Ted Sumers

(website) Language is the bedrock of human society, yet despite intensive study its role in cognition remains mysterious. My research combines reinforcement learning and language games to explore human communication in complex decision-making settings. In particular, I'm using formalisms from reinforcement learning to develop rational models of joint action. I'm applying these insights to both advance our understanding of uniquely human cognition (e.g., how language supports cultural evolution) and develop artificial intelligence which can interact successfully with people (e.g., improving natural language interfaces for value alignment).


Xuechunzi Bai

Xuechunzi Bai

(website) Broadly, I am interested in applying computational methods and formal models to classic social psychological ideas. Currently, I am interested in two topics; both of them investigate the collateral damage from an otherwise functional approach. In one line of research, I examine how inaccurate stereotypes can result from rational explorations. In another line of research, I explore how social inequality can emerge from rational resource transmissions.


Lab Manager

Maya Malaviya

Maya Malaviya

Maya is fascinated by curiosity and learning. She hopes to use recent findings from psychology, neuroscience, and computer science to improve pedagogical choices and learning environments. She completed her B.A. in Cognitive Science at University of California, Berkeley. Maya also works in Tania Lombrozo's Concepts and Cognition Lab.


Alumni and Long-Distance Affiliates

Joshua Abbott

Joshua Abbott

Joe Austerweil

Joe Austerweil

Vincent Berthiaume

Vincent Berthiaume

Wesley Baraff Bonawitz

Liz Bonawitz

David Bourgin

David Bourgin

Daphna Buchsbaum

Daphna Buchsbaum

Kevin Canini

Kevin Canini

Daniel Chada

Daniel Chada

Dawn Chen

Dawn Chen

Ishita Dasgupta

Ishita Dasgupta

Naomi Feldman

Naomi Feldman

Vael Gates

Vael Gates

Sharon Goldwater

Sharon Goldwater

Jessica Hamrick

Jessica Hamrick

Chris Holdgraf

Chris Holdgraf

Anne Hsu

Anne Hsu

Tiffany Hwu

Tiffany Hwu

Nori Jacoby

Nori Jacoby

Rachel Jansen

Rachel Jansen

Peaks Krafft

Peaks Krafft

Casey Lewry

Casey Lewry

Falk Lieder

Falk Lieder

Chris Lucas

Chris Lucas

Aida Nematzadeh

Aida Nematzadeh

Stephan Meylan

Stephan Meylan

Luke Maurits

Luke Maurits

Thomas Morgan

Thomas Morgan

M Pacer

M Pacer

Alexandra Paxton

Alexandra Paxton

Jay Martin

Jay Martin

Avi Press

Avi Press

Anna Rafferty

Anna Rafferty

Florencia Reali

Florencia Reali

Daniel Reichman

Daniel Reichman

Adam Sanborn

Adam Sanborn

Sophia Sanborn

Sophia Sanborn

Benj Shapiro

Benj Shapiro

Lei Shi

Lei Shi

Jordan Suchow

Jordan Suchow

Bill Thompson

Bill Thompson

Andrew Whalen

Andrew Whalen

Joseph Jay Williams

Joseph Jay Williams

Frank Wood

Frank Wood

Jing Xu

Jing Xu

Saiwing Yeung

Saiwing Yeung

Julia Ying

Julia Ying

Qiong Zhang

Qiong Zhang

© 2022 Computational Cognitive Science Lab  |  Department of Psychology  |  Princeton University