Towards Neural Functional Program Evaluation
This work addresses program evaluation for functional programming languages, offering incremental improvements in neural methods for this specific domain.
The paper investigates transformer-based language models for evaluating functional programs, introducing a program generation method to control syntactic sugar, and finds that T5 models achieve over 90% exact match scores on in-distribution and out-of-distribution tests, with pretrained weights outperforming random initialization.
This paper explores the capabilities of current transformer-based language models for program evaluation of simple functional programming languages. We introduce a new program generation mechanism that allows control over syntactic sugar for semantically equivalent programs. T5 experiments reveal that neural functional program evaluation performs surprisingly well, achieving high 90% exact program match scores for most in-distribution and out-of-distribution tests. Using pretrained T5 weights has significant advantages over random initialization. We present and evaluate on three datasets to study generalization abilities that are specific to functional programs based on: type, function composition, and reduction steps. Code and data are publicly available at https://github.com/ElementAI/neural-interpreters.