Learning Program Synthesis for Integer Sequences from Scratch
DOI:
https://doi.org/10.1609/aaai.v37i6.25930Keywords:
ML: Unsupervised & Self-Supervised Learning, ML: Reinforcement Learning Algorithms, ML: Transparent, Interpretable, Explainable ML, SNLP: Language ModelsAbstract
We present a self-learning approach for synthesizing programs from integer sequences. Our method relies on a tree search guided by a learned policy. Our system is tested on the On-Line Encyclopedia of Integer Sequences. There, it discovers, on its own, solutions for 27987 sequences starting from basic operators and without human-written training examples.
Downloads
Published
2023-06-26
How to Cite
Gauthier, T., & Urban, J. (2023). Learning Program Synthesis for Integer Sequences from Scratch. Proceedings of the AAAI Conference on Artificial Intelligence, 37(6), 7670-7677. https://doi.org/10.1609/aaai.v37i6.25930
Issue
Section
AAAI Technical Track on Machine Learning I
