Exploring Software Naturalness through Neural Language Models

The Software Naturalness hypothesis argues that programming languages can be understood through the same techniques used in natural language processing. We explore this hypothesis through the use of a pre-trained transformer-based language model to perform code analysis tasks... (read more)

Results in Papers With Code
(↓ scroll down to see all results)