Alien Coding
Thibault Gauthier, Miroslav Ol\v{s}\'ak, Josef Urban

TL;DR
This paper presents a self-learning algorithm that autonomously synthesizes programs for thousands of OEIS sequences by combining neural machine translation with iterative program proposal, discovering novel methods.
Contribution
It introduces a novel self-learning approach that combines neural translation and program synthesis to automate OEIS sequence programming.
Findings
Discovered programs for over 78,000 OEIS sequences
Developed unusual programming methods autonomously
Analyzed the behavior and creativity of the algorithm
Abstract
We introduce a self-learning algorithm for synthesizing programs for OEIS sequences. The algorithm starts from scratch initially generating programs at random. Then it runs many iterations of a self-learning loop that interleaves (i) training neural machine translation to learn the correspondence between sequences and the programs discovered so far, and (ii) proposing many new programs for each OEIS sequence by the trained neural machine translator. The algorithm discovers on its own programs for more than 78000 OEIS sequences, sometimes developing unusual programming methods. We analyze its behavior and the invented programs in several experiments.
Peer Reviews
No public reviews on file for this paper yet. If you reviewed it on a platform where reviews are public (OpenReview, ICLR, NeurIPS, ICML), you can paste yours below so the community can read it here.
Videos
No videos yet. Explain this paper in a talk, walkthrough, or lecture? Add one.
Taxonomy
TopicsNeural Networks and Applications · Evolutionary Algorithms and Applications
MethodsSelf-Learning
