Mental Representations as Distribution-Sensitive Data Structures.

Zenna Tavares, Armando Solar-LezamaAAAI 2017 · OpenReview

Abstract

We model representations as data-structures which are distribution sensitive, i.e., which exploit regularities in their usage patterns to reduce time or space complexity. We introduce probabilistic axiomatic specifications to extend abstract data structures - which specify a class of representations with equivalent logical behavior - to a distribution-sensitive data structures. We reformulate synthesis of distribution-sensitive data structures as a continuous function approximation problem, using deep neural networks to induce stacks, queues, and natural number representations.
BibTeX
@inproceedings{tavares2017mental,
  author = {Zenna Tavares and Armando Solar-Lezama},
  title = {{Mental Representations as Distribution-Sensitive Data Structures.}},
  booktitle = {AAAI Conference on Artificial Intelligence},
  year = {2017},
  url = {https://openreview.net/forum?id=HpBQpXeShfml},
}

Loading related content...

Figure