To What Extent do Deep Learning-based Code Recommenders Generate Predictions by Cloning Code from the Training Set? — mentioned by belter

Sun, 17 Apr 2022 20:38:32 GMT

"...Deep Learning (DL) models have been widely used to support code completion. These models, once properly trained, can take as input an incomplete code component (e.g., an incomplete function) and predict the missing tokens to finalize it. GitHub Copilot is an example of code recommender built by training a DL model on millions of open source repositories... ...As of Today, it is unclear whether the code generated by DL models trained on open source code should be considered as “new” or as “derivative” work, with possible implications on license infringements. In this work, we run a…

Paper and discussion history. A mention is not an endorsement.

To What Extent do Deep Learning-based Code Recommenders Generate Predictions by Cloning Code from the Training Set? — HN submission

Sun, 17 Apr 2022 20:38:31 GMT

To What Extent Deep Learning Code Recommenders Generate Predictions by Cloning?

1 point; 2 discussion replies on HN. Paper and source history. A submission is not an endorsement.

To What Extent do Deep Learning-based Code Recommenders Generate Predictions by Cloning Code from the Training Set? · HackerPapers discussions

To What Extent do Deep Learning-based Code Recommenders Generate Predictions by Cloning Code from the Training Set? — mentioned by belter

To What Extent do Deep Learning-based Code Recommenders Generate Predictions by Cloning Code from the Training Set? — HN submission