Neural Functional Programming

John K. Feser, Marc Brockschmidt, Alexander L. Gaunt, Daniel Tarlow

TL;DR: A differentiable functional programming language for learning programs from input-output examples.

Abstract: We discuss a range of modeling choices that arise when constructing an end-to-end differentiable programming language suitable for learning programs from input-output examples. Taking cues from programming languages research, we study the effect of memory allocation schemes, immutable data, type systems, and built-in control-flow structures on the success rate of learning algorithms. We build a range of models leading up to a simple differentiable functional programming language. Our empirical evaluation shows that this language allows to learn far more programs than existing baselines.

Keywords: Supervised Learning

Conflicts: microsoft.com, csail.mit.edu