Combining Small Language Models and Large Language Models for Zero-Shot NL2SQL

Summary: ZeroNL2SQL: fine-tuned SLMs (with novel database serialization and question-aware schema alignment) generate SQL sketches, then LLMs fill missing clauses and values. Multi-level value matching plus execution-based selection yields SOTA zero-shot NL2SQL (+5.5–20% execution accuracy). (summarized by gpt-5-mini on Feb 09 2026)

Paper ID: 13684
Venue: VLDB
Year: 2024
Pagerank: 7.1249098e-05
Overall Rank: 3,787 | 74.02%
DOI: 10.14778/3681954.3681960

Incoming Non-self Citations Over Time

Authors

1. Ju Fan (Renmin University of China)
2. Zihui Gu (Renmin University of China)
3. Songyue Zhang (Renmin University of China)
4. Yuxin Zhang (Renmin University of China)
5. Zui Chen (Massachusetts Institute of Technology)
6. Lei Cao (Massachusetts Institute of Technology; University of Arizona)
7. Guoliang Li (Tsinghua University)
8. Samuel Madden (Massachusetts Institute of Technology)
9. Xiaoyong Du (Renmin University of China)
10. Nan Tang (Hong Kong University of Science and Technology)

BibTeX Citation

@article{fan_vldb24,
        title = {{Combining Small Language Models and Large Language Models for Zero-Shot NL2SQL}},
        author = {Fan, Ju and Gu, Zihui and Zhang, Songyue and Zhang, Yuxin and Chen, Zui and Cao, Lei and Li, Guoliang and Madden, Samuel and Du, Xiaoyong and Tang, Nan},
        journal = {PVLDB},
        series = {{VLDB} '24},
        volume = {17},
        number = {11},
        pages = {2753--2763},
        doi = {10.14778/3681954.3681960},
        url = {https://doi.org/10.14778/3681954.3681960},
        year = {2024}
}

Incoming Citations (Sorted by Pagerank)

Showing 19 of 19 citing papers.

Rank	Citing Paper	Year	Venue	Pagerank
6,254	In-depth Analysis of Graph-based RAG in a Unified Framework	2025	VLDB	5.9405174e-05
6,259	R-Bot: An LLM-based Query Rewrite System	2025	VLDB	5.93967e-05
8,627	ThriftLLM: On Cost-Effective Selection of Large Language Models for Classification Queries	2025	VLDB	5.3969543e-05
10,253	HAMMER: An Automatic RAG Tuning System via Hierarchical Memory-Guided Monte Carlo Tree Search	2026	SIGMOD	5.093636e-05
10,276	OctoSelector: Efficient and Effective Batch-Aware Model Selection for Large Language Models	2026	SIGMOD	5.093636e-05
10,285	Reward-SQL: Boosting Text-to-SQL via Stepwise Execution-Aware Reasoning and Process-Supervised Rewards	2026	SIGMOD	5.093636e-05
10,304	VecBench: A Controllable Benchmark for Filtered Vector Search: [Experiments & Analysis]	2026	SIGMOD	5.093636e-05
10,340	AgentTune: An Agent-Based Large Language Model Framework for Database Knob Tuning	2026	SIGMOD	5.093636e-05
10,344	Are Your LLM-based Text-to-SQL Models Secure? Exploring SQL Injection via Backdoor Attacks	2026	SIGMOD	5.093636e-05
10,374	GeoKGM: A Multimodal Large Language Model for Zero-Shot Knowledge Graph Completion in Geospatial Databases	2026	SIGMOD	5.093636e-05
10,389	PLForge: Enhancing Language Models for Natural Language to Procedural Extensions of SQL	2026	SIGMOD	5.093636e-05
10,398	Reliable Answers for Recurring Questions: Boosting Text-to-SQL Accuracy with Template Constrained Decoding	2026	SIGMOD	5.093636e-05
10,505	The Case For Language Model Approximated LIKE Predicate	2026	SIGMOD	5.093636e-05
10,506	This is Going to Sound Crazy, But What If We Used Large Language Models to Boost Automatic Database Tuning Algorithms By Leveraging Prior History? We Will Find Better Configurations More Quickly Than Retraining From Scratch!	2026	SIGMOD	5.093636e-05
10,510	NL2SQLBench: A Modular Benchmarking Framework for LLM-Enabled NL2SQL Solutions	2026	VLDB	5.093636e-05
10,537	TACO: A Benchmark for Open-Domain Text-to-SQL with Ambiguous and Cross-Database Queries	2026	VLDB	5.093636e-05
10,556	OpenSQL: Data-Efficient Text-to-SQL for Open-Source LLMs via Synthesized Intermediate Supervision	2026	VLDB	5.093636e-05
10,867	Weak-to-Strong Prompts with Lightweight-to-Powerful LLMs for High-Accuracy, Low-Cost, and Explainable Data Transformation	2025	VLDB	5.093636e-05
11,013	Towards Automated Cross-domain Exploratory Data Analysis through Large Language Models	2025	VLDB	5.093636e-05

Outgoing Citations (Sorted by Pagerank)

Showing 7 of 7 cited papers.

Citations counted here include only citations to other VLDB/SIGMOD/CIDR/PODS papers in this database.

Rank	Cited Paper	Year	Venue	Pagerank
279	Text-to-SQL Empowered by Large Language Models: A Benchmark Evaluation	2024	VLDB	0.00022468369
756	CodeS: Towards Building Open-source Language Models for Text-to-SQL	2024	SIGMOD	0.0001431656
1,550	Symphony: Towards Natural Language Query Answering over Multi-modal Data Lakes	2023	CIDR	0.00010385904
2,748	Few-shot Text-to-SQL Translation using Structure and Content Prompt Learning	2023	SIGMOD	8.1707811e-05
3,436	Unicorn: A Unified Multi-tasking Model for Supporting Matching Tasks in Data Integration	2023	SIGMOD	7.4157897e-05
5,730	Domain Adaptation for Deep Entity Resolution	2022	SIGMOD	6.1071585e-05
11,545	OpenTFV: An Open Domain Table-Based Fact Verification System	2022	SIGMOD	5.093636e-05

Semantically Similar Papers

#	Overall Rank	Paper	Year	Venue
1	279	Text-to-SQL Empowered by Large Language Models: A Benchmark Evaluation	2024	VLDB
2	2,748	Few-shot Text-to-SQL Translation using Structure and Content Prompt Learning	2023	SIGMOD
3	6,936	Reliable Text-to-SQL with Adaptive Abstention	2025	SIGMOD
4	865	Natural language to SQL: Where are we today?	2020	VLDB
5	1,444	CatSQL: Towards Real World Natural Language to SQL Applications	2023	VLDB
6	9,305	The Power of Constraints in Natural Language to SQL Translation	2025	VLDB
7	2,852	The Dawn of Natural Language to SQL: Are We Fully Ready?	2024	VLDB
8	2,602	NL2SQL is a solved problem... Not!	2024	CIDR
9	9,870	Natural Language to SQL: State of the Art and Open Problems	2025	VLDB
10	10,510	NL2SQLBench: A Modular Benchmarking Framework for LLM-Enabled NL2SQL Solutions	2026	VLDB