Storage Capacity of Kernel Associative Memories

By Barbara Caputo, Heinrich Niemann. 2002

23 citations recorded by the source corpus.

Subfields: Artificial Intelligence, Cognitive Neuroscience, Statistical and Nonlinear Physics, Model Architectures

How researchers see this paper

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

In Artificial Intelligence

1. Günter Klambauer
2. Hubert Ramsauer
3. Sepp Hochreiter
4. Angela Bitto-Nemling
5. Markus Holzleitner
6. Hamid Eghbalzadeh
7. Fabian Paischer
8. Fei Tang

In Cognitive Neuroscience

1. Hava T. Siegelmann
2. Sepp Hochreiter
3. Dimitri Nowicki

In Statistical and Nonlinear Physics

1. Angela Bitto-Nemling
2. Hubert Ramsauer
3. Viet-Hung Tran
4. Sepp Hochreiter

In Model Architectures

1. Günter Klambauer
2. Angela Bitto-Nemling
3. Sepp Hochreiter
4. Hubert Ramsauer
5. Markus Holzleitner
6. Hamid Eghbalzadeh
7. Fabian Paischer
8. Vihang Patil

Citations in the corpus

• CLOOB: Modern Hopfield Networks with InfoLOOB Outperform CLIP (2021, background): “Their storage capacity can be considerably increased by polynomial terms in the energy function (Chen et al., 1986; Psaltis & Cheol, 1986; Baldi & Venkatesh, 1987; Gardner, 1987; Abbott & Arian, 1987; Horn & Usher, 1988; Caputo & Niemann, 2002; Krotov & Hopfield, 2016).”
• History Compression via Language Models in Reinforcement Learning (2022, background): “Their storage capacity can be considerably increased by polynomial terms in the energy function (Chen et al., 1987; Psaltis & Park, 1986; Baldi & Venkatesh, 1987; Gardner, 1987; Abbott & Arian, 1987; Horn & Usher, 1988; Caputo & Niemann, 2002; Krotov & Hopfield, 2016).”
• Flexible Kernel Memory (2010, background): “Caputo [41] explored analogies from associative memory to ‘‘kernel spin glass’’ and demonstrated an attractor network, loading bipolar inputs and using generalized Hebbian learning to load non-flexible memories with greater memory capacity than the Hopfield network.”