In-ide code generation from natural language: Promise and challenges

FF Xu, B Vasilescu, G Neubig - ACM Transactions on Software …, 2022 - dl.acm.org
… for natural language to code (NL2Code) generation and retrieval as in-IDE developer …
assigned to complete a range of Python programming tasks with and without the use of the two …

Productivity assessment of neural code completion

A Ziegler, E Kalliamvakou, XA Li, A Rice… - Proceedings of the 6th …, 2022 - dl.acm.org
… Another study considered the benefits of using a plugin converting natural language prompts
to code [18]. It found no statistically significant improvements in task completion time or task …

Towards Advancing Code Generation with Large Language Models: A Research Roadmap

H Jin, H Chen, Q Lu, L Zhu - arXiv preprint arXiv:2501.11354, 2025 - arxiv.org
generating code from descriptions of natural language and validating it against the test sets
[1, 11, 26, 28]. Other key tasks include code … explore the challenges in code generation from …

Predictive synthesis of API-centric code

D Nam, B Ray, S Kim, X Qu, S Chandra - Proceedings of the 6th ACM …, 2022 - dl.acm.org
… 15] or natural language description [30… This is challenging as we go to a larger number
of API functions, because we cannot cover all permutations exhaustively. However, the model …

[ОПИСАНИЕ][C] In-ide code generation from natural language: Promise and challenges. corr, abs/2101.11149 (2021)

FF Xu, B Vasilescu, G Neubig - arXiv preprint arXiv:2101.11149, 2021