Learning Continuous Attractors in Recurrent Networks

NeurIPS 1997 pp. 654-660

/neurips/1997/seung1997neurips-learning/

Abstract

One approach to invariant object recognition employs a recurrent neu(cid:173) ral network as an associative memory. In the standard depiction of the network's state space, memories of objects are stored as attractive fixed points of the dynamics. I argue for a modification of this picture: if an object has a continuous family of instantiations, it should be represented by a continuous attractor. This idea is illustrated with a network that learns to complete patterns. To perform the task of filling in missing in(cid:173) formation, the network develops a continuous attractor that models the manifold from which the patterns are drawn. From a statistical view(cid:173) point, the pattern completion task allows a formulation of unsupervised learning in terms of regression rather than density estimation.

PDF NeurIPS Semantic Scholar

Cite

Text

Seung. "Learning Continuous Attractors in Recurrent Networks." Neural Information Processing Systems, 1997.

Markdown

[Seung. "Learning Continuous Attractors in Recurrent Networks." Neural Information Processing Systems, 1997.](https://mlanthology.org/neurips/1997/seung1997neurips-learning/)

BibTeX

@inproceedings{seung1997neurips-learning,
  title     = {{Learning Continuous Attractors in Recurrent Networks}},
  author    = {Seung, H. Sebastian},
  booktitle = {Neural Information Processing Systems},
  year      = {1997},
  pages     = {654-660},
  url       = {https://mlanthology.org/neurips/1997/seung1997neurips-learning/}
}