Dataflow Matrix Machines and Vvalues: a Bridge between Programs and Neural Nets
This topic contains 0 replies, has 1 voice, and was last updated by arXiv 11 months, 4 weeks ago.

Dataflow Matrix Machines and Vvalues: a Bridge between Programs and Neural Nets
Dataflow matrix machines generalize neural nets by replacing streams of numbers with streams of vectors (or other kinds of linear streams admitting a notion of linear combination of several streams) and adding a few more changes on top of that, namely arbitrary input and output arities for activation functions, countablesized networks with finite dynamically changeable active part capable of unbounded growth, and a very expressive selfreferential mechanism. While recurrent neural networks are Turingcomplete, they form an esoteric programming platform, not conductive for practical generalpurpose programming. Dataflow matrix machines are more suitable as a generalpurpose programming platform, although it remains to be seen whether this platform can be made fully competitive with more traditional programming platforms currently in use. At the same time, dataflow matrix machines retain the key property of recurrent neural networks: programs are expressed via matrices of real numbers, and continuous changes to those matrices produce arbitrarily small variations in the programs associated with those matrices. Spaces of vectorlike elements are of particular importance in this context. In particular, we focus on the vector space $V$ of finite linear combinations of strings, which can be also understood as the vector space of finite prefix trees with numerical leaves, the vector space of “mixed rank tensors”, or the vector space of recurrent maps. This space, and a family of spaces of vectorlike elements derived from it, are sufficiently expressive to cover all cases of interest we are currently aware of, and allow a compact and streamlined version of dataflow matrix machines based on a single space of vectorlike elements and variadic neurons. We call elements of these spaces Vvalues. Their role in our context is somewhat similar to the role of Sexpressions in Lisp.
Dataflow Matrix Machines and Vvalues: a Bridge between Programs and Neural Nets
by Michael Bukatin, Jon Anthony
https://arxiv.org/pdf/1712.07447v1.pdf
You must be logged in to reply to this topic.