|
|
|
|
||
|
|
|
 |
|
|
|
|||||||||||||||||||||||||||||||||||||||||||||||
|
|
|
Seminar Series on Theoretical Neuroscience This spring, the Redwood Neuroscience Institute, in collaboration with Stanford University, will host four public lectures on the Stanford campus by leading thinkers in theoretical neuroscience. Despite decades of research and general progress in experimental neuroscience, we still don't have an overall theory of how our brains work. In this series we have asked the speakers to share their perspective on this problem and to give their outlook on the development of a brain theory. The first speaker is Christoph von der Malsburg (April 10) from the Ruhr-University Bochum, Germany and USC. In the early 1980s, von der Malsburg proposed what still remains a highly controversial theory---namely, that neural representations of objects in the environment, although spatially separate within the brain, are bound together by synchronous firing patterns. Some evidence has accumulated for this theory over the past decade, but many scientists are still skeptical, arguing that the rate of firing alone constitutes the basis of neural representation. Some theorists have gone so far as to claim that synchronous firing could also form the neural substrate of conscious experience. One of the proponents of this idea, Christof Koch (April 24), from the California Institute of Technology, will be the next speaker in the series. In the early 1990s, Koch teamed together with Francis Crick to propose that consciousness should be studied seriously as a topic of neuroscience, rather than purely in the domain of philosophy. They point out, for example, that anesthesiologists make their living by temporarily knocking out consciousness, yet no one really understands how. Thus, they argue, it is incumbent upon neuroscientists to investigate and better understand the neural basis of consciousness, rather than dismiss it as "too obscure." Koch now collaborates with a neurosurgery group at UCLA investigating the neural substrate of awareness in humans. The final two speakers in the series---Geoff Hinton (May 15) of the University of Toronto, and Terry Sejnowski (June 12) of the Salk Institute---are best known for their work on learning in neural networks. In the mid-1980s, they were instrumental in showing how synaptic weights in a neural network could be automatically adjusted, through experience, in order to train the network in certain tasks. In recent years, it has been shown how these learning rules can account for the response properties of neurons at early stages of sensory processing. Hinton is currently exploring learning in hierarchical neural-network models that emulate the hierarchical architecture of the cortex. Sejnowski is currently exploring the role of thalamocortical interactions during sleep, thought to be crucial for the consolidation of memories. NOTE: For abstracts of seminars, click on the title of the talk.
|
|
||||||||||||||||||||||||||||||||||||||||||||||
 |
||||||||||||||||||||||||||||||||||||||||||||||||||
|
|
|
© 2002 Redwood Institute. All rights reserved. |
|
||||||||||||||||||||||||||||||||||||||||||||||