1 paper
Jiahui Chen, Lei Fu, Jian Cui +2
Large-scale Text-to-SQL benchmarks such as BIRD typically assume complete and accurate database annotations as well as readily available external knowledge, which fails to reflect…