Brain Day

09:15

Psychology: Jeffrey D. Schall
Professor of Biology
Canada Research Chair in Translating Neuroscience
Director, Visual Neurophysiology Center
York University

TITLE: 
Neuro-Computational Investigations of Decision Making in Visual Search

ABSTRACT: 
The first part of this presentation will summarize previous neurophysiological studies in macaque monkeys doing visual search tasks. The findings of these studies (subsequently replicated by many others) distinguished neurons representing the salience (or priority) of an item relative to others and other neurons that generated gaze shifts based on the location and features of the salient item. The relationship between these two types of neural representation were then articulated through two cognitive computational models. First, the Salience by Competitive and Recurrent Interactions (SCRI) model explains how neurons derive salience from location and identification representations. The model rationalized the observed heterogeneity of the neurons. Second, as an instance of a canonical stochastic accumulator decision process, the Gated Accumulator model explains how the neurons generating gaze shifts can be understood as accumulating the salience evidence. With this background, the second part of this presentation will introduce new research we have been doing that has a rich history in human cognitive psychology but had never been used with monkeys. We trained monkeys to perform a visual search task with factorial manipulations of two factors: singleton-distractor identifiability and stimulus-response cue discriminability. I will summarize a mathematical theory (systems factorial technology) we use to distinguish processing architectures from patterns of response times, and I will forecast new neurophysiological findings that offer confirmation of the original distinction between salience and response preparation operations but demonstrate the inadequacy of the canonical accumulator model of decision making.

10:45 

Computation:  Robert E. Kass
Maurice Falk University Professor of Statistics and Computational Neuroscience,
Department of Statistics & Data Science,
Machine Learning Department, and 
Neuroscience Institute
Carnegie Mellon University

TITLE:
Identification of Interacting Neural Populations: Methods and Statistical Issues

ABSTRACT: 
I will discuss the primary statistics-in-neuroscience topic my trainees and I have worked on over the past eight years: it is the problem of identifying interactions among neural populations from large-scale electrophysiological recordings. I will describe statistically natural approaches we have used to document interactions among brain areas based on neural spike trains (using latent variable point process models) and oscillating field potentials (by defining an exponential family on a multidimensional torus). I will also highlight some of the high-level lessons I've learned about the nature of statistical inference in science, which suggest ways we can continue to improve the scientific process. 

Neuroscience:  Michale S. Fee 
Glenn V. and Phyllis F. Dorflinger Professor of Neuroscience
Department Head, Department of Brain and Cognitive Sciences,
Investigator, McGovern Institute for Brain Research
Massachusetts Institute of Technology

TITLE:
 A Neural Clock Underlying the Temporal Structure of an Auditory Memory

ABSTRACT:  
Songbirds, such as zebra finches, store an auditory memory of their tutor's song and, over the course of several months during learning, gradually refine their own vocalizations to produce a close match to the tutor song. A young zebra finch can learn to imitate a song it hears only a few times, even though their songs are temporally complex and are extremely precisely reproduced. How is such a precise auditory memory of the tutor song formed, and how is it recalled during vocal practice? It is known that adult song is controlled by a sparse neuronal sequence in the pre-motor nucleus HVC. The observation that local cooling of HVC slows the song suggests that timing is controlled by local dynamics within HVC. However, recent work also shows that disruption of HVC or its auditory inputs and outputs during tutoring impairs vocal learning, suggesting a role for HVC in tutor memory formation and recall. Here we propose a specific model by which neural dynamics in HVC form during tutoring and act as a precise neural clock, not just for singing, but also for laying down the auditory tutor memory and synchronizing the recall of the tutor memory during singing, allowing the temporally precise readout of vocal errors. Motivated by this hypothesis, we tested a critical prediction of our theory: that cooling of HVC during tutoring leads to a sped-up tutor song memory, analogous to slowing the motor of a tape recorder during recording and then playing the recording back at normal speed. We designed a modular thermoelectric cooling device and found that transient cooling during tutoring caused birds to produce faster imitations, consistent with our theory. This work gives insight into formation of auditory memories, and how they can be stably recalled later to guide precise sensory-motor learning.

Philosophy: John Heil
​Professor of Philosophy
Department of Philosophy
Washington University in St. Luis,

TITLE: 
Take the Explanatory Gap. Please.

ABSTRACT: 
Few philosophers nowadays doubt the existence and significance of a persistent ‘explanatory gap’ in our understanding of the nature of conscious experiences and their relation to the material world. Concerns about the explanatory gap have their roots in Saul Kripke’s 1972 argument against the mind-brain identity theory: if a is identical with b, then there is no world at which a fails to be identical with b; as Descartes showed, however, it is conceivable for minds to exist in the absence of material bodies; so, Kripke concluded, minds cannot be identified with material bodies. In 1983 Joseph Levine argued that, although Kripke’s original argument falls short of establishing that minds are distinct from material bodies, the argument has an epistemological counterpart. The disparate character of conscious qualities and qualities of material bodies creates an ineliminable barrier to our understanding how the mental could be identified with the physical. This, and other, expressly epistemological arguments have subsequently been deployed – illicitly – in the service of the metaphysical thesis originally defended by Kripke.