RESEARCHInvestigateNEXT 12 MONTHS
PLSQLBench: Benchmarking LLM Systems for Executable Procedural Database Programming
arXiv cs.CL — Computation and Language
Factual evidence
What the source reports
Researchers introduced PLSQLBench, a 2,865-instance execution-based benchmark evaluating LLM capabilities in procedural PL/SQL code generation.
Open source