From: Anders Sandberg (asa@nada.kth.se)
Date: Wed Jun 11 2003 - 03:42:34 MDT
On Mon, Jun 09, 2003 at 07:13:28AM -0700, Robert J. Bradbury wrote:
>
> Interesting stats that at least in monkeys it takes a dozen trials
> to establish a memory and that 2-4 new associations were formed
> during a recording session (presumably perhaps a few hours).
I think we humans are rather extreme in our fast learning. I predict that
we will discover both finely tuned neuromodulatory control of memory
formation and that we have extremely fast learning in some parts of the
brain.
> Also makes me wonder -- if the hippocampus is responsible for memory then
> what the heck is the rest of the brain doing?
As Ramez said, it is likely consolidation. We use the hippocampus and
related systems for initial storage, and then it is moved to the cortex
(likely duering sleep or rest). People with hippocampal damage still
recall things learned prior to the damage, and can learn procedural
memories despite being unable to learn new episodic memories.
The real issue IMHO is how the transfer of memory works; in our neural
network research we have a loss rate of 50% of memory patterns, which
makes me wonder if we either store memories redundantly in the
hippocampus or if some hi-fi transfer occurs. Or if we actually do have
this awful learning but don't notice it.
-- ----------------------------------------------------------------------- Anders Sandberg Towards Ascension! asa@nada.kth.se http://www.nada.kth.se/~asa/ GCS/M/S/O d++ -p+ c++++ !l u+ e++ m++ s+/+ n--- h+/* f+ g+ w++ t+ r+ !y
This archive was generated by hypermail 2.1.5 : Wed Jun 11 2003 - 03:50:46 MDT