Videos from Cognitive Computational Neuroscience (CCN) 2017

Wednesday, September 6, 2017

Welcome

Welcome: Why are we meeting?
Thomas Naselaris, Medical University of South Carolina (MUSC)

Kavli Opening Keynote Session: Explaining Cognition, Brain Computation, and Intelligent Behavior

Cognitive Science:

The cognitive science perspective: Reverse-engineering the mind
Josh Tenenbaum, Massachusetts Institute of Technology

Computational Neuroscience:
Nicole Rust, University of Pennsylvania

Artificial Intelligence:

How does the brain learn so much so quickly?
Yann LeCun, Facebook, New York University

Panel Discussion:

Nancy Kanwisher, Massachusetts Institute of Technology
Josh Tenenbaum, Massachusetts Institute of Technology
Nicole Rust, University of Pennsylvania
Yann LeCun, Facebook, New York University
Bruno Olshausen, UC-Berkeley
Jackie Gottlieb, Columbia University
Moderator: Jim DiCarlo, Massachusetts Institute of Technology

Contributed Talks: Sensation & Perception

Neural networks for efficient bayesian decoding of natural images from retinal neurons
Eleanor Batty, Columbia University

A reverse correlation test of computational models of lightness perception
Richard Murray, York University

Keynote 1

I have not decided yet
Michael N. Shadlen, Columbia University

Contributed Talks: Attention & Judgement

Understanding biological visual attention using convolutional neural networks
Grace Lindsay, Columbia University

A dichotomy of visual relations or the limits of convolutional neural networks
Matthew Ricci & Junkyung Kim, Brown University

Computation of human confidence reports in decision-making with multiple alternatives
Hsin-Hung Li, New York University

Thursday, September 7, 2017

Tutorial talks

Cognitive Science:

Modeling of behavior
Wei Ji Ma, New York University

Computational Neuroscience:

Tutorial on computational neuroscience
Ruben Coen-Cagli, Albert Einstein College of Medicine

Artificial Intelligence:

Artificial Intelligence with connections to neuroimaging
Alona Fyshe, University of Victoria

Contributed Talks: Memorability & predictive coding

Unconscious perception of scenes reveals a perceptual neural signature of memorability
Yalda Mohsenzadeh, Massachusetts Institute of Technology

Predictive coding & neural communication delays produce alpha-band oscillatory impulse response functions
Rufin VanRullen, Université Toulouse

Keynote 2

How we understand others’ emotions
Rebecca Saxe, Massachusetts Institute of Technology

Contributed Talks: Localization & task learning

Emergence of grid-like representations by training recurrent neural networks to perform spatial localization
Chris Cueva, Columbia University

Modular task learning in an embodied two-dimensional visual environment
Kevin T. Feigelis, Stanford University

Distributed mechanisms supporting information search and value-based choice in prefrontal cortex
Laurence Hunt, University of Oxford

Friday, September 8, 2017

Keynote 3

Probabilistic models of sensorimotor control
Daniel Wolpert, University of Cambridge

Contributed Talks: Reinforcement learning & control

Surprise, surprise: Cholinergic and dopaminergic neurons encode complementary forms of reward prediction errors
Fitz Sturgill, Cold Spring Harbor Laboratory

Hippocampal pattern separation supports reinforcement learning
Ian Ballard, Stanford University

Offline replay supports planning: fMRI evidence from reward revaluation
Ida Momennejad, Princeton University

Keynote 4

How we learn and represent the structure of tasks
Yael Niv, Princeton University

Contributed Talks: Exploration & exploitation

Amygdala drives value and exploration signals in striatum and orbitofrontal Cortex
Vincent Costa, National Institute of Mental Health

History effect in a minimalistic explore-exploit task
Mingyu Song, Princeton University

Keynote 5

Strategic decision-making in the human subcortex measured with UHF-MRI
Birte Forstmann, University of Amsterdam

Contributed Talks: Learning in deep neural networks

Deep learning with segregated dendrites
Blake Aaron Richards, University of Toronto Scarborough

When do neural networks learn sequential solution in short-term memory tasks?
Emin Orhan, New York University

Closing Keynote Debate: What is the best level to model the mind-brain?

Perspective 1: Cognitive Models:

Bridging the computational and algorithmic levels
Tom Griffiths, UC-Berkeley

Perspective 2: Deep Learning and the Cognitive Brain:

Deep learning and backprop in the brain
Yoshua Bengio, Université de Montréal

Panel Discussion:

Tom Griffiths, UC-Berkeley
Yoshua Bengio, Université de Montréal
Anne Churchland, Cold Spring Harbor Laboratory
Aude Oliva, Massachusetts Institute of Technology
Tony Movshon, New York University
Moderator:
Jonathan Cohen, Princeton University

Closing

Closing Remarks
Nikolaus Kriegeskorte, Columbia University

NYC Video Production by Dreambox Productions