Data-aware candidate selection in NL2SQL translation via small separating instances
数据库
2026-05-13 v1
摘要
We propose a data-aware candidate selection method for NL2SQL translation based on separating instances and provenance. We implement this approach and evaluate it against three natural baselines on a subset of BIRD-DEV. Experiments show that our method significantly outperforms baselines when only two or three candidates are given and no consistency score is available. The code of our prototype can be found at https://github.com/staskikotx/SISelection
引用
@article{arxiv.2605.12319,
title = {Data-aware candidate selection in NL2SQL translation via small separating instances},
author = {Stanislav Kikot and Alexander Shulgin and Yanwei Xu},
journal= {arXiv preprint arXiv:2605.12319},
year = {2026}
}
备注
8 pages