Neurons with graded response have collective computational properties like those of two-state neurons.

By John J. Hopfield. 1984

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Natural Language Processing

Citations in the corpus

CLOOB: Modern Hopfield Networks with InfoLOOB Outperform CLIP (2021, background): “Hopﬁeld networks are energy-based, binary associative memories, which popularized artiﬁcial neural networks in the 1980s (Amari, 1972; Hopﬁeld, 1982, 1984).”
A neurobiological model of visual attention and invariant pattern recognition based on dynamic routing of information (1993, methodology): “[The effect of adding this term is discussed in Hopfield (1984).”
A Constructive Mean-Field Analysis of Multi-Population Neural Networks with Random Synaptic Weights and Stochastic Inputs (2008, neutral): “The voltage-based model The assumption, made in [25], is that the post-synaptic potential has the same shape no matter which presynaptic population caused it, the sign and amplitude may vary though.”
Toward an Integration of Deep Learning and Neuroscience (2016, background): “BPTT “unfolds” a recurrent dynamics (Hopﬁeld, 1984) in which hidden units are perturbed toward target values (Bengio and Fischer, 2015), loosely similar to that which occurs in XCAL.”
Self-Avoiding Random Dynamics on Integer Complex Systems (2011, methodology): “Ising models, also known as Boltzmann machines, are ubiquitous models in physics, machine learning and spatial statistics [Ackley et al. 1985; Besag 1974; Hopfield 1984; Kindermann and Snell 1980; Newman and Barkema 1999].”
Generalization in Interactive Networks: The Benefits of Inhibitory Competition and Hebbian Learning (2001, background): “Computationally, interactivity enables iterative, constraint-satisfaction-style processing (Smolensky, 1986; Ackley, Hinton, & Sejnowski, 1985; Hopeeld, 1982, 1984 ), and generally allows information processing to eow in many different directions within one network, providing a richer and more eexi”
Portfolio selection using neural networks (2005, methodology): “The Hopfield network [13] is an artificial NN model with a single layer of neurons fully connected.”