Is This You, LLM? How to Recognize AI-written Multilingual Programs with Reproducible Code Stylometry - Replication package and CLI tool

Is This You, LLM? How to Recognize AI-written Multilingual Programs with Reproducible Code Stylometry This is the replication package for the "Is This You, LLM? How to Recognize AI-written Multilingual Programs with Reproducible Code Stylometry" paper and a CLI tool for human-AI authorship. The repository contains two main experiments from the paper: - The data mining process - The authorship attribution models training, here divided into two steps (trainer_single_provenance and trainer_multi_provenance) The package also contains: - Checkpoints for both of the multi-language models - The dataset *H_AIRosettaMP.csv* - A CLI tool with a brief utilization guide in the README.md file Further information regarding the replication of the experiments is in the README.md file.