User profiles for Samuel Gershman
Samuel GershmanProfessor, Harvard University Verified email at fas.harvard.edu Cited by 22703 |
Building machines that learn and think like people
Recent progress in artificial intelligence has renewed interest in building systems that learn
and think like people. Many advances have come from using deep neural networks trained …
and think like people. Many advances have come from using deep neural networks trained …
Reinforcement learning and episodic memory in humans and animals: an integrative framework
SJ Gershman, ND Daw - Annual review of psychology, 2017 - annualreviews.org
We review the psychology and neuroscience of reinforcement learning (RL), which has
experienced significant progress in the past two decades, enabled by the comprehensive …
experienced significant progress in the past two decades, enabled by the comprehensive …
Computational rationality: A converging paradigm for intelligence in brains, minds, and machines
After growing up together, and mostly growing apart in the second half of the 20th century,
the fields of artificial intelligence (AI), cognitive science, and neuroscience are reconverging …
the fields of artificial intelligence (AI), cognitive science, and neuroscience are reconverging …
[HTML][HTML] Model-based influences on humans' choices and striatal prediction errors
The mesostriatal dopamine system is prominently implicated in model-free reinforcement
learning, with fMRI BOLD signals in ventral striatum notably covarying with model-free …
learning, with fMRI BOLD signals in ventral striatum notably covarying with model-free …
A tutorial on Bayesian nonparametric models
SJ Gershman, DM Blei - Journal of Mathematical Psychology, 2012 - Elsevier
A key problem in statistical modeling is model selection, that is, how to choose a model at an
appropriate level of complexity. This problem appears in many settings, most prominently in …
appropriate level of complexity. This problem appears in many settings, most prominently in …
The hippocampus as a predictive map
A cognitive map has long been the dominant metaphor for hippocampal function, embracing
the idea that place cells encode a geometric representation of space. However, evidence …
the idea that place cells encode a geometric representation of space. However, evidence …
Reinforcement learning in multidimensional environments relies on attention mechanisms
In recent years, ideas from the computational field of reinforcement learning have revolutionized
the study of learning in the brain, famously providing new, precise theories of how …
the study of learning in the brain, famously providing new, precise theories of how …
The curse of planning: dissecting multiple reinforcement-learning systems by taxing the central executive
A number of accounts of human and animal behavior posit the operation of parallel and
competing valuation systems in the control of choice behavior. In these accounts, a flexible but …
competing valuation systems in the control of choice behavior. In these accounts, a flexible but …
Retrospective revaluation in sequential decision making: a tale of two systems.
Recent computational theories of decision making in humans and animals have portrayed 2
systems locked in a battle for control of behavior. One system—variously termed model-free …
systems locked in a battle for control of behavior. One system—variously termed model-free …
The successor representation: its computational logic and neural substrates
SJ Gershman - Journal of Neuroscience, 2018 - Soc Neuroscience
Reinforcement learning is the process by which an agent learns to predict long-term future
reward. We now understand a great deal about the brain's reinforcement learning algorithms, …
reward. We now understand a great deal about the brain's reinforcement learning algorithms, …