Publications

Probabilistic models of physical reasoning

PDF

Integrating heuristic and simulation-based reasoning in intuitive physics

The ability to predict, reason about, and act in the physical world is crucial for human survival, but the cognitive systems that …
PDF

"Just In Time" representations for mental simulation in intuitive physics

Many models of intuitive physical reasoning posit some kind of mental simulation mechanism, yet everyday environments frequently …
PDF

Physion++: Evaluating physical scene understanding that requires online inference of different physical properties

General physical scene understanding requires more than simply localizing and recognizing objects – it requires knowledge that objects …
PDF Project

Are deep neural networks SMARTer than second graders?

Recent times have witnessed an increasing number of applications of deep neural networks towards solving tasks that require superior …
PDF

An approximate representation of objects underlies physical reasoning

People make fast and reasonable predictions about the physical behavior of everyday objects. To do so, people may be using principled …
PDF

H-SAUR: Hypothesize, Simulate, Act, Update, and Repeat for understanding object articulations from interactions

The world is filled with articulated objects that are difficult to determine how to use from vision alone, e.g., a door might open …
PDF

Neural evidence that intuitive physics guides visual tracking and working memory

Our perceptions are rooted in early-developing expectations about the behavior of everyday objects, which are often exposed by …
PDF

Mental Jenga: A counterfactual simulation model of physical support

From building towers to picking an orange from a stack of fruit, assessing support is critical for successfully interacting with the …
PDF

Partial mental simulation explains fallacies in physical reasoning

People can reason intuitively, efficiently, and accurately about everyday physical events. Recent accounts suggest that people use …
PDF

Reconsidering the 'bias' in 'the correspondence bias'

We do not directly observe the internal qualities of others so we must infer them from behavior. Although classic attribution theories …
PDF

AGENT: A benchmark for core psychological reasoning

For machine agents to successfully interact with humans in real-world settings, they will need to develop an understanding of human …
PDF Project

Meta-strategy learning in physical problem-solving: The effect of embodied experience

‘Embodied cognition’ suggests that our experience in our bodies – including our motor experiences – shape our cognitive and perceptual …
PDF

Physion: Evaluating physical prediction from vision in humans and machines

While machine learning algorithms excel at many challenging visual tasks, it is unclear that they can make predictions about …
PDF

Unsupervised discovery of 3D physical objects from video

We study the problem of unsupervised physical object discovery. Unlike existing frameworks that aim to learn to decompose scenes into …
PDF

Data-efficient learning for complex and real-time physical problem solving using augmented simulation

Humans quickly solve tasks in novel systems with complex dynamics, without requiring much interaction. While deep reinforcement …
PDF

Rapid trial-and-error learning with simulation supports flexible tool use and physical reasoning

Many animals, and an increasing number of artificial agents, display sophisticated capabilities to perceive and manipulate objects. But …
PDF Project Video

The fine structure of surprise in intuitive physics: when, why, and how much?

We are surprised when events violate our intuitive physical expectations. Even infants look longer when things seem to magically …
PDF Poster Video

Abstract strategy learning underlies flexible transfer in physical problem solving

What do people learn when they repeatedly try to solve a set of related problems? In a set of three different exploratory physical …
PDF

Modeling expectation violation in intuitive physics with coarse probabilistic object representations

From infancy, humans have expectations about how objects will move and interact. Even young children expect objects not to move through …
PDF Code Project

End-to-end differentiable physics for learning and control

We present a differentiable physics engine that can be integrated as a module in deep neural networks for end-to-end learning. As a …
PDF

Neurocomputational modeling of human physical scene understanding

Human scene understanding involves not just localizing objects, but also inferring latent attributes that affect how the scene might …
PDF Poster

Learning to act by integrating mental simulation and physical experiments

People can learn about the effects of their actions either by performing physical experiments or by running mental simulations. …
PDF

Different physical intuitions exist between tasks, not domains

Does human behavior exploit deep and accurate knowledge about how the world works, or does it rely on shallow and often inaccurate …
PDF

Tiptoeing around it: Inference from absence in potentially offensive speech

Language that describes people in a concise manner may con- flict with social norms (e.g., referring to people by their race), …
PDF

Differentiable physics and stable modes for tool-use and manipulation planning

We consider the problem of sequential manipulation and tool-use planning in domains that include physical interactions such as hitting …
PDF

Thinking inside the box: Motion prediction in contained spaces uses simulation

Theories of the mental processes people use to perform physical reasoning often differ on whether they are based on simulation or on …
PDF Poster

Faulty towers: A hypothetical simulation model of physical support

In this paper we introduce the hypothetical simulation model (HSM) of physical support. The HSM predicts that people judge physical …
PDF

Think again? The amount of mental simulation tracks uncertainty in the outcome

In this paper, we investigate how people use mental simulations: do people vary the number of simulations that they run in order to …
PDF

The "Fundamental Attribution Error" is rational in an uncertain world

Others’ internal qualities (e.g. dispositions, attitudes) are not directly observable so we must infer them from behavior. Classic …
PDF

Prospective uncertainty: The range of possible futures in physical prediction

Recent research has suggested that people make physical predictions based on extrapolation from a noisy representation of the world, …
PDF Poster

The role of sequential dependence in creative semantic search

PDF

Looking forwards and backwards: Similarities and differences in prediction and retrodiction

People must often infer what might have transpired in the past to bring about the present state of the world, a task called …
PDF

Empirical evidence for Markov Chain Monte Carlo in memory search

Previous theoretical work has proposed the use of Markov chain Monte Carlo as a model of exploratory search in memory. In the current …
PDF

Reductionism and practicality

Like most domains of science, the study of the mind has been tackled at many scales of analysis, from the behavior of large groups of …
PDF

Physical predictions over time

In order to interact with the world, people must be able to predict how it will unfold in the future, and these predictions must be …
PDF

Consistent physics underlying ballistic motion prediction

Research into human models of intuitive physics typically falls into one of two camps, either claiming that intuitive physics is biased …
PDF Poster

Multiply-constrained semantic search in the Remote Associates Task

Many important problems require consideration of multiple constraints, such as choosing a job based on salary, location, and …
PDF

Put your money where your mouth is: Incentivizing the truth my making nonreplicability costly

We argue that every published result should be backed by an author-issued ‘nonreplication bounty’: an amount of money the author is …
PDF

Sources of uncertainty in intuitive physics

Recent work suggests that people predict how objects interact in a manner consistent with Newtonian physics, but with additional …
PDF