papers on consolidation of long-term memories

[Meeter M]

Apologies for multiple postings.

Dear Colleagues,

The readers of this list may be interested in the following four papers on
the consolidation of long-term memory. Of all papers a draft is available at
www.neuromod.org/publications. 


* Meeter, M. & Murre, J.M.J. (in press). TraceLink: A model of amnesia and
consolidation. Cognitive Neuropsychology.

A connectionist model is presented, the TraceLink model, that implements an
autonomous 'off-line' consolidation process. The model consists of three
subsystems: (1) a trace system (neocortex), (2) a link system (hippocampus
and adjacent regions), and (3) a modulatory system (basal forebrain and
other areas). The model is able to account for many of the characteristics
of anterograde and retrograde amnesia, including Ribot gradients, transient
global amnesia, patterns of shrinkage of retrograde amnesia, correlations
between anterograde and retrograde amnesia or the absence thereof (e.g., in
isolated retrograde amnesia). In addition, it produces normal forgetting
curves and can exhibit permastore. It also offers an explanation for the
advantages of learning under high arousal for long-term retention.
http://www.neuromod.org/publications/papers/TL/TL.pdf


* Meeter, M. & Murre, J.M.J. (2004). Simulating episodic memory deficits in
semantic dementia with the TraceLink model. Memory, 12, 272 - 287.

Although semantic dementia is primarily characterized by deficits in
semantic memory, episodic memory is also impaired. Patients show poor recall
of old autobiographical and semantic memories, with better retrieval of
recent experiences; they can form new memories, and normal performance on
pictorial recognition memory has been demonstrated. As these abnormalities
in episodic memory are virtually a mirror image of those seen in the amnesic
syndromes, semantic dementia poses a challenge to extant models of remote
memory and amnesia. Here, we show that one such model, TraceLink, can
reproduce some of the principal findings on episodic memory in semantic
dementia. A loss of nodes and connections within the trace system, which can
be identified with the temporal neocortical memory storage sites implicated
in semantic dementia, simulates without further assumptions the findings
reported above.
http://www.neuromod.org/publications/papers/SEMD/SEMD.pdf


* Meeter, M. (2003). Control of consolidation in neural networks: Avoiding
runaway effects. Connection Science, 15, 45-61.

Consolidation has been implemented in two ways: as straight rehearsal of
patterns, or as pseudorehearsal, in which pseudoitems are created by
sampling attractors or input-output combinations from the network. Although
several authors have investigated both implementations, few have explored
how it is decided which pattern or pseudoitem is consolidated. Controlling
consolidation is not trivial, as it is susceptible to a corruption. In
runaway consolidation, one or two patterns monopolize all consolidation
resources and come to dominate the entire network. Runaway consolidation is
analysed here, and three solutions are explored. Suppressing transmission in
the connections in which consolidation takes place is shown to work best.
Placing bounds on connections or unlearning attractors also alleviates
runaway consolidation, though less effectively so.
http://www.neuromod.org/publications/papers/CCNN/CCNN.pdf


* Meeter, M. & Murre, J.M.J. (in press). Consolidation of long-term memory:
Evidence and alternatives. Psychological Bulletin.

Memory loss in retrograde amnesia has long been held to be larger for recent
periods than for remote periods, a pattern usually referred to as the Ribot
gradient. One explanation for this gradient is consolidation of long-term
memories. Several computational models of such a process have been
presented, and have shown how consolidation can explain characteristics of
amnesia, effects of arousal on memory, and episodic memory in semantic
dementia. These models have not elucidated how consolidation must be
envisaged; it remains a largely hypothetical process. Here findings are
reviewed that shed light on how consolidation may be implemented in the
brain. Moreover, evidence is weighed that supports its existence, or
supports one of its competitors as explanations of the Ribot gradient: that
it results from only partial damage to the hippocampal memory system (as
stated by multiple trace theory), or that it result from a sparing of
semanticized memories. Consolidation theory, multiple trace theory and
semantization can all handle some findings well, and others not. Conclusive
evidence for or against consolidation thus remains to be found.
http://www.neuromod.org/publications/papers/CLM/CLM.pdf


Sincerely,
Martijn Meeter


------------------------------
 Dr. M. Meeter
 Dept. of Cognitive Psychology
 Vrije Universiteit Amsterdam
 The Netherlands
------------------------------