concept

active

concept:storing-infinite-numbers-of-patterns-in-a-spin-glass-model-of-neural-networks-amit-et-al-1985

Storing infinite numbers of patterns in a spin-glass model of neural networks (Amit et al., 1985)

Result that original Hopfield network memory capacity scales linearly with network dimensionality; background for scaling discussion.

Neighborhood — ranked by edge-count

Papers (1)

paper

Relating transformers to models and neural representations of the hippocampal formation
cites

Concepts (1)

concept

Hopfield Network Memory Capacity Scaling
cites
Technical issue: original Hopfield networks scale linearly with neuron count; exponential activations enable 2^(N/2) scaling but softmax used in TEM-t has intermediate properties.

Related by similarity (8)

cosine ≥ 0.65 · no typed edge

Entities in the same semantic neighborhood but without a typed relation to this one — candidates for new edges or unrecognized duplicates.

Hopfield network can store multiple patterns, recall them via content-addressable memory, and generalise to novel patterns with the same underlying structure.finding0.774
Summary of known Hopfield network capabilities used as a model for collective computation.
Neural networks, trained with different objectives on different data and modalities, are converging to a shared statistical model of reality in their representation spaces.quote0.772
The paper's central thesis statement, presented prominently after the abstract
On a Model of Associative Memory with Huge Storage Capacity (Demircigil et al., 2017)concept0.762
Analytical result showing exponential power activation allows memory storage scaling as 2^(N/2); cited in context of Hopfield scaling.
Different neural network models trained on different objectives and modalities are converging to a shared statistical model of reality in their representation spaceshypothesis0.752
The central hypothesis of the paper; the platonic representation hypothesis itself
Patterns capture knowledge beyond the individual.claim0.749
Why technologists love Alexander; patterns as mechanism for sharing and reusing design knowledge.
Bioelectric pattern memories store target morphology for anatomical homeostasis.claim0.746
Planarian head number can be permanently altered by re-writing bioelectric prepatterns.
The ability to represent relationships (e.g. correlations or associations) among variables, rather than a system of independent variables, is essential for any connectionist model of cognition.claim0.738
Necessary condition for connectionist cognition.
Neural Representations of Location Composed of Spatially Periodic Bands (Krupic et al., 2012)concept0.736
Discovery of band cells; TEM-t also recapitulates these representations.