Holographic Recurrent Networks

Abstract

Holographic Recurrent Networks (HRNs) are recurrent networks which incorporate associative memory techniques for storing se(cid:173) quential structure. HRNs can be easily and quickly trained using gradient descent techniques to generate sequences of discrete out(cid:173) puts and trajectories through continuous spaee. The performance of HRNs is found to be superior to that of ordinary recurrent net(cid:173) works on these sequence generation tasks.

Cite

Text

Plate. "Holographic Recurrent Networks." Neural Information Processing Systems, 1992.

Markdown

[Plate. "Holographic Recurrent Networks." Neural Information Processing Systems, 1992.](https://mlanthology.org/neurips/1992/plate1992neurips-holographic/)

BibTeX

@inproceedings{plate1992neurips-holographic,
  title     = {{Holographic Recurrent Networks}},
  author    = {Plate, Tony A.},
  booktitle = {Neural Information Processing Systems},
  year      = {1992},
  pages     = {34-41},
  url       = {https://mlanthology.org/neurips/1992/plate1992neurips-holographic/}
}