Content area
Full Text
A R T I C L E S
Computational principles of synaptic memory consolidation
Marcus K Benna1 & Stefano Fusi1,2
Memories are stored and retained through complex, coupled processes operating on multiple timescales. To understand the computational principles behind these intricate networks of interactions, we construct a broad class of synaptic models that efficiently harness biological complexity to preserve numerous memories by protecting them against the adverse effects of overwriting. The memory capacity scales almost linearly with the number of synapses, which is a substantial improvement over the square root scaling of previous models. This was achieved by combining multiple dynamical processes that initially store memories in fast variables and then progressively transfer them to slower variables. Notably, the interactions between fast and slow variables are bidirectional. The proposed models are robust to parameter perturbations and can explain several properties of biological memory, including delayed expression of synaptic modifications, metaplasticity, and spacing effects.
2016 Nature America, Inc., part of Springer Nature. All rights reserved.
The complexity and diversity of the numerous biological mechanisms that underlie memory is both fascinating and disconcerting. The molecular machinery responsible for memory consolidation at the level of synaptic connections is believed to employ a complex network of diverse biochemical processes that operate on different timescales1,2. Understanding how these processes are orchestrated to preserve memories over a lifetime requires guiding principles to interpret the complex organization of the observed synaptic molecular interactions and explain its computational advantage. Here we present a class of synaptic models that can efficiently harness biological complexity to store and preserve a huge number of memories on long timescales, vastly outperforming all previous synaptic models of memory.
The models we construct solve a long-standing dilemma: on the one hand, in a memory system that is continually receiving and storing new information, synaptic strengths representing memories must be protected from being overwritten during the storage of new information. Failure to provide such protection results in memory lifetimes that are catastrophically low35. On the other hand, protecting old memories too rigidly causes memory traces of new information to be weak, being represented by small numbers of synapses. This is one aspect of the plasticityrigidity dilemma69. Synapses that are highly plastic are good at storing new memories but poor at...