A sparse quantized hopfield network for online-continual memory

Nature Communications, Год журнала: 2024, Номер 15(1)

Опубликована: Май 2, 2024

An important difference between brains and deep neural networks is the way they learn. Nervous systems learn online where a stream of noisy data points are presented in non-independent, identically distributed way. Further, synaptic plasticity brain depends only on information local to synapses. Deep networks, other hand, typically use non-local learning algorithms trained an offline, non-noisy, independent, setting. Understanding how under same constraints as open problem for neuroscience neuromorphic computing. A standard approach this has yet be established. In paper, we propose that discrete graphical models via maximum posteriori algorithm could provide such approach. We implement kind model network called Sparse Quantized Hopfield Network. show our outperforms state-of-the-art associative memory tasks, these online, continual settings, learns efficiently with inputs, better than baselines episodic task.

Язык: Английский

---

Learning probability distributions of sensory inputs with Monte Carlo predictive coding

PLoS Computational Biology, Год журнала: 2024, Номер 20(10), С. e1012532 - e1012532

Опубликована: Окт. 30, 2024

It has been suggested that the brain employs probabilistic generative models to optimally interpret sensory information. This hypothesis formalised in distinct frameworks, focusing on explaining separate phenomena. On one hand, classic predictive coding theory proposed how can be learned by networks of neurons employing local synaptic plasticity. other neural sampling theories have demonstrated stochastic dynamics enable circuits represent posterior distributions latent states environment. These frameworks were brought together variational filtering introduced coding. Here, we consider a variant for static inputs, which refer as Monte Carlo (MCPC). We demonstrate integration with results network learns precise using computation and The MCPC infer presence generate likely inputs their absence. Furthermore, captures experimental observations variability activity during perceptual tasks. By combining sampling, account both sets data previously had explained these individual frameworks.

Язык: Английский

---

High-order rotor Hopfield neural networks for associative memory

Neurocomputing, Год журнала: 2024, Номер unknown, С. 128893 - 128893

Опубликована: Ноя. 1, 2024

Язык: Английский

---

Learning Sequence Attractors in Recurrent Networks with Hidden Neurons

Опубликована: Янв. 1, 2024

Download This Paper Open PDF in Browser Add to My Library Share: Permalink Using these links will ensure access this page indefinitely Copy URL DOI

Язык: Английский

---

Learning sequence attractors in recurrent networks with hidden neurons

Neural Networks, Год журнала: 2024, Номер 178, С. 106466 - 106466

Опубликована: Июнь 22, 2024

Язык: Английский

---