REVIEW 3 cited by
Random Features Hopfield Networks generalize retrieval to previously unseen examples
Not yet reviewed by Pith; the record is open.
This paper has not been read by Pith yet. Machine review is queued; the pith claim, tier, and objections will appear here once it completes.
SPECIMEN: schema-true, not a live event
T0 review · schema-true
One-sentence machine reading of the paper's core claim.
pith:XXXXXXXX · record.json · timestamp
read the original abstract
It has been recently shown that a learning transition happens when a Hopfield Network stores examples generated as superpositions of random features, where new attractors corresponding to such features appear in the model. In this work we reveal that the network also develops attractors corresponding to previously unseen examples generated with the same set of features. We explain this surprising behaviour in terms of spurious states of the learned features: we argue that, increasing the number of stored examples beyond the learning transition, the model also learns to mix the features to represent both stored and previously unseen examples. We support this claim with the computation of the phase diagram of the model.
Forward citations
Cited by 3 Pith papers
-
Exponential Capacity in Multilayer Hetero-Associative Neural Networks
A multilayer exponential Hopfield network stores e^{Nρ_L} hetero-associative patterns, with ρ_L∼L log2 and basins that match simulated, immune-receptor, and language data.
-
Implicit Bias and Invariance: How Hopfield Networks Efficiently Learn Graph Orbits
Hopfield networks can memorize entire graph isomorphism classes with polynomially many samples, aided by an implicit norm-minimization bias that drives weights toward a 3-dimensional invariant subspace.
-
Hopfield Networks as Models of Emergent Function in Biology
A review article that explains Hopfield network mathematics and interprets the dynamics as signal retrieval, subspace projection, and energy landscape descent, then surveys applications to cell fate, self-assembly, an...
Discussion (0). Continue with ORCID to comment.