Hasson, Uri Princeton
<p>age approx: 42 or so; BS 94 frrom Hebrew Univ Jersalem;
PhD neurobio from the Weizmann;
to 2008 NYU David Heeger;
from 2008 Asst Prof Princeton.</p>
<p>
Feb 28, 2014 02:39 PM</p>
<p>
FriSem today abstract:</p>
<p>
As we perceive and comprehend the nested information in our environment, our brains must integrate past information with present input across multiple timescales. For example, a word only achieves full meaning in the context of a sentence, and a sentence only achieves full meaning in the context of a passage of text. The capacity to actively accumulate and integrate information over multiple timescales is, therefore, crucial to the functional processes of most brain areas. However, traditional models of short term memory (STM) and working memory (WM) propose dedicated STM and WM memory modules (buffers) which are separate from the neural circuits that process incoming information. In this talk I will outline an alternative model, arguing that there are no dedicated buffers of working memory. In contrast the model argues that memory is a distributed property of neural circuits, and that each circuit can actively accumulate information over time as part of its information processing role. Moreover, in order to integrate real-life information over multiple timescales, active memory processing timescales are organized hierarchically, so that the integration window increases from early sensory areas to higher-order perceptual and cognitive areas. Thus, memories of the recent past are not stored in a few localized working memory buffers, but rather are hierarchically distributed throughout the nervous system.</p>
<p>
jkjkj</p>