Pervasive Annotation Errors Break Text-to-SQL Benchmarks and Leaderboards
Summary: Empirical audit of text-to-SQL benchmark annotations: BIRD Mini-Dev and Spider 2.0-Snow show massive error rates (52.8%/62.8%). Correcting BIRD Dev changes agent scores by -7% to +31% and leaderboard ranks by up to 9 positions, showing benchmarks can mislead model selection. (summarized by gpt-5.4-mini on Apr 12 2026)
Incoming Non-self Citations Over Time
No non-self incoming citations found for this paper in this database.
Authors
- 1. Tengjun Jin
- 2. Yoojin Choi
- 3. Yuxuan Zhu
- 4. Daniel Kang
Incoming Citations (Sorted by Pagerank)
Showing 0 of 0 citing papers.
| Rank | Citing Paper | Year | Venue | Pagerank |
|---|
Previous
Page 1 / 1
Next
Outgoing Citations (Sorted by Pagerank)
Showing 4 of 4 cited papers.
Citations counted here include only citations to other VLDB/SIGMOD/CIDR/PODS papers in this database.
| Rank | Cited Paper | Year | Venue | Pagerank |
|---|---|---|---|---|
| 369 | Text-to-SQL Empowered by Large Language Models: A Benchmark Evaluation | 2024 | VLDB | 0.0002547515 |
| 998 | CodeS: Towards Building Open-source Language Models for Text-to-SQL | 2024 | SIGMOD | 0.00014729379 |
| 3,859 | OpenSearch-SQL: Enhancing Text-to-SQL with Dynamic Few-shot and Consistency Alignment | 2025 | SIGMOD | 6.6907933e-05 |
| 3,978 | OmniSQL: Synthesizing High-quality Text-to-SQL Data at Scale | 2025 | VLDB | 6.5725884e-05 |
Previous
Page 1 / 1
Next