Pith. sign in

REVIEW 3 major objections 6 minor 40 references

TailorSQL: An NL2SQL System Tailored to Your Query Workload

T0 review · 3 major / 6 minor · reviewed 2026-08-07 · deepseek-v4-flash

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗

Pith's one-line read Workload-specific retrieval built from past SQL logs improves NL2SQL execution accuracy by up to 2× and uses up to 15× fewer prompt tokens at matched accuracy.

desk verdict TailorSQL is a competent, genuinely novel workload-aware NL2SQL system, but its headline accuracy gains are measured against baselines deprived of PK-FK metadata that real databases usually have. read the letter →

arxiv 2505.23039 v1 pith:TYU6GJU2 submitted 2025-05-29 cs.DB cs.CL

Kapil Vaidya , Jialin Ding , Sebastian Kosak , David Kernert , Chuan Lei , Xiao Qin , Abhinav Tripathy , Ramesh Balan
show 2 more authors
Balakrishnan Narayanaswamy Tim Kraska
This is my paper
classification cs.DBcs.CL
keywords NL2SQLtext-to-SQLretrieval-augmentedgenerationqueryworkloadspecializationdocumentretrievalexecutionaccuracyabstentionpolicy
verification ladder T0 review T1 audit T2 compute T3 formal

The pith

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The reading

TailorSQL tries to establish that the SQL queries a database has already executed are a reusable resource for NL2SQL, not just the schema itself. It claims that by mining past logs for common join paths, filter expressions, and group-by clauses, and by using those logs to shape both document embeddings and per-class prompt-token budgets, a retrieval-augmented text-to-SQL pipeline can generate more accurate SQL while consuming less context. On three benchmarks the paper reports execution-accuracy gains of 12.5, 10.9, and 22.7 percentage points over the strongest retrieval baseline, and matched accuracy with 1.8×, 15×, and 1.9× fewer prompt tokens. If this is right, existing databases carry transferable semantic signal—what cryptic tables mean, which joins are conventional, how column values are formatted—that schema-only retrieval misses.

What carries the argument

The central object is the workload-tailored document embedding $E_{\mathrm{doc}} = w_1 E_{\mathrm{raw}} + w_2 E_{\mathrm{co\text{-}occur}} + w_3 E_{\mathrm{SQL}} + w_4 E_{\mathrm{synthQ}}$, with weights optimized once per workload by gradient descent on a cosine-loss objective over a synthetic question workload generated from past queries. It is paired with a context allocator that uses Bayesian optimization to split the prompt token budget across table, column, and hint document classes, and a bandit-based abstention policy that switches to a generic schema-only pipeline when user feedback indicates the specialized pipeline no longer matches the workload.

What would settle it

Run TailorSQL on a production database with a time-ordered split: train the specialized embeddings and context allocation on the first several months of real query logs and evaluate on the next several months of real user questions, comparing against schema-only retrieval with the same LLM. If execution accuracy on genuinely future questions does not beat the baseline, the central claim that past logs transfer to future questions fails.

Watch

Extended reading notes

Core claim

The paper's central discovery is that the historical query workload contains reusable semantic structure for NL2SQL, and that this structure can be injected into a retrieval-augmented generation pipeline at three points: document content, document embeddings, and retrieval budgets. Concretely, past queries are broken into clause-level hint documents covering join paths, filters, and group-by conditions; each document's embedding is reweighted as a blend of its raw embedding, embeddings of co-occurring documents, embeddings of the SQL queries that used it, and embeddings of synthetic user questions generated from those queries; and a Bayesian optimizer chooses how many prompt tokens to give to table, column, and hint documents. Evaluated as execution accuracy, TailorSQL reports consistent gains over workload-agnostic retrieval, with the largest gains on the benchmark with the most cryptic table names and complex nested queries.

Load-bearing premise

The load-bearing premise is that LLM-generated synthetic questions derived from past SQL queries faithfully resemble the natural-language questions real users will ask later; if those synthetic questions are unrepresentative, the tailored embeddings and context allocation are tuned to the wrong signal.

Editorial extensions

If this is right

  • Deployed NL2SQL systems on existing databases can improve accuracy without new data collection: the query logs are already there.
  • At matched execution accuracy, the prompt can be 1.8×–15× smaller, which directly lowers per-query LLM latency and cost.
  • Workload specialization composes with existing reasoning-based NL2SQL pipelines, so it can be added on top of current state-of-the-art systems rather than replacing them.
  • The abstention policy gives a safe deployment story: when the workload shifts, the system can fall back to generic retrieval instead of silently degrading.
  • Hint documents for joins, filters, and group-bys carry most of the value, so the scheme is extensible to other SQL clause types.

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

  • Editorial inference: TailorSQL's Random Split evaluation likely overstates production gains; a time-ordered split, where training logs precede test questions, would be a stricter test of whether past logs transfer to genuinely future questions.
  • Editorial inference: the realism of LLM-generated synthetic questions is untested; having human annotators rate or rewrite a sample and measuring how retrieval accuracy changes would expose how much the method depends on that link.
  • Editorial inference: the same embedding-blend recipe could transfer to other retrieval-augmented settings where historical artifacts encode conventions—e.g., code generation from repository commit logs—but this is outside the paper's evidence.
  • Editorial inference: because fine-tuned embeddings achieved higher recall than TailorSQL's weighted blend, a hybrid that periodically fine-tunes a small per-workload adapter could combine TailorSQL's low offline cost with higher retrieval recall.
Share X Bluesky LinkedIn Reddit HN

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, and a circularity audit.

Referee Report

3 major / 6 minor

Summary. The paper proposes TailorSQL, a retrieval-augmented NL2SQL system that specializes to a database's past query workload. Offline, it builds schema documents and query-hint documents (join-path, filter, and group-by hints) from historical SQL logs, then computes document embeddings as a weighted sum of raw SBERT embeddings and proxy embeddings derived from co-occurring documents, past SQL queries, and LLM-generated synthetic questions; the weights are optimized on the logged workload. It also uses Bayesian optimization to allocate the LLM prompt context among document classes. At runtime, TailorSQL retrieves documents per class, builds a prompt, and generates SQL; a bandit-based abstention policy decides when to fall back to a generic schema-only pipeline. The evaluation on three benchmark-derived "Union" databases reports 12.5%, 10.9%, and 22.7% higher execution accuracy than the best baseline, with substantially fewer prompt tokens at matched accuracy, and shows that TailorSQL can improve existing systems such as DIN-SQL and MAC-SQL.

Significance. If the reported results hold, the core idea is valuable: past query logs contain transferable semantic information, such as common join paths and column semantics, that schema-only RAG misses. The paper's system design is coherent, the ablations isolate the main components, and using the same LLM across all compared systems is a good control. The abstention-policy idea is a sensible safeguard against workload drift. However, the headline quantitative claims currently rest on an evaluation setup that removes standard schema metadata (PK-FK constraints) while preserving log-derived join hints, and on a random split that makes the test workload similar to the logged workload by construction. The synthetic questions used as the training signal are not validated. These are load-bearing concerns for the central claim that workload specialization, rather than restoration of routine metadata or benchmark distribution overlap, drives the reported gains.

major comments (3)
  1. [§8.1, footnote 2; §8.2.2] The experimental setup excludes primary-key/foreign-key (PK-FK) information from the schema, while TailorSQL's hint documents include join-path hints extracted from the past query log. In real database catalogs, FK constraints are routinely available; the SBERT and BM25 baselines are therefore deprived of information that TailorSQL effectively restores through its join-path hints, and group-by hints can similarly substitute for missing primary keys. The 12.5%, 10.9%, and 22.7% accuracy advantages over the best baseline in §8.3 may thus reflect restoration of standard catalog metadata rather than novel information from past queries. To support the paper's central claim, the authors should rerun the comparison with PK-FK constraints included in the schema documents for all systems, or at least report baseline accuracy with intact PK-FK metadata.
  2. [§8.1, §8.3, §8.5] The headline end-to-end accuracy is reported under the Random Split, where the logged and test halves are drawn from the same benchmark distribution. The paper itself calls this setup 'ideal for showcasing the benefits of TailorSQL' (§8.1), but a real deployment may face a log that only partially overlaps the future workload. The Disjoint Split is used only to evaluate the abstention policy (§8.5), not to report TailorSQL's end-to-end accuracy against the baselines. Without an evaluation under a less favorable split (for example, BIRD train versus dev, or a temporal split), the claim that TailorSQL specializes to real workloads rather than to the benchmark distribution is not established.
  3. [§5.1, Eq. (2), §6] The embedding weights and the context allocation are optimized on synthetic questions generated by an LLM from the logged SQL queries, but the paper does not validate the fidelity of these synthetic questions to real user phrasing. There is no report of the generation prompt, the number of synthetic questions, or any comparison of their distribution with real user questions. If the synthetic questions are systematically more aligned with the logged SQL than real user paraphrases are, the learned weights in Eq. (2) and the Bayesian-optimization objective in §6 could be overfit to the synthetic workload, inflating the reported accuracy. Please add a validation study of synthetic-question quality (for example, human judgments, or an ablation that uses real logged questions instead of synthetic ones) and disclose the generation details.
minor comments (6)
  1. [§1, §2] The term 'query' is used both for natural-language user questions and for SQL queries, which can be confusing in Section 5 where 'past queries' are SQL and 'user questions' are natural language; consistent terminology would improve readability.
  2. [§8.1] The construction of Bird-Union and Spider-Union (combining all dev-set databases into one) is not described in detail; it is unclear whether same-named tables or columns across original databases were renamed and how this affects PK-FK relationships, so the schema-removal decision is hard to reproduce.
  3. [§8.3] The paper reports median values over five runs but does not report variance or significance for the end-to-end accuracy differences; given the modest dataset sizes (300-1200 questions) and possible split variance, confidence intervals would strengthen the comparisons.
  4. [§7, §8.5] The epsilon value and sliding-window length used in the bandit-based abstention policy are not reported, which makes the workload-drift experiment difficult to reproduce.
  5. [Table 1] The end-to-end ablation is reported only on Bird-Union; reporting the same ablations on at least one other benchmark would strengthen the claim that each component is necessary.
  6. [§8.2.2] The paper does not report the number of hint documents generated per benchmark or the distribution across the three hint types, which would help assess the relative contribution of each hint type to the accuracy gains.

Circularity Check

0 steps flagged · score 0.0 of 10

No circularity: workload specializations are fit on the logged half and evaluated on a held-out half; the main caveats are evaluation-design issues, not circular reduction.

full rationale

TailorSQL's claimed prediction—higher execution accuracy from workload specialization—is not equivalent to its inputs by construction. The embedding mixture weights (Eq. 2), the context allocation, and the hint documents are all computed offline from the query-log half of each benchmark split, and the reported accuracy in Section 8.3 is measured on the held-out half (Random Split or Disjoint Split). Test questions and their ground-truth SQL are not used to fit the embedding weights, the token allocations, or the hint contents, so the accuracy numbers are a genuine held-out evaluation rather than a renamed training objective. The generated synthetic questions are an auxiliary training signal, not the reported prediction target. The self-citations present in the paper—[30] for long SQL text and [25] as the FIBEN benchmark—are not load-bearing: the first is corroborated by an external reference and the second is a dataset, not a uniqueness theorem or ansatz smuggled in to force the conclusion. The footnote in Section 8.1 that removes PK-FK information from the schema while join-path hints from the query log remain is a legitimate evaluation-fairness and external-validity caveat, but it does not make the derivation circular: both TailorSQL and the baselines operate in the same modified schema, and there is no equation or fitted parameter that forces the reported 12.5%, 10.9%, and 22.7% gaps. Concerns about synthetic-question realism and about Random Split sharing distribution with the log are correctness-risk issues, not circular reductions. No step in the derivation chain reduces to its own input.

Assumptions & free parameters 3 free parameters · 4 assumptions · 0 invented entities

No speculative physical entities are introduced. The system uses software artifacts such as hint documents and synthetic questions, which are not invented entities in the sense of new particles or forces. The trained parameters are fit on the training split of the workload and not on the test questions, so they are not circular in the narrow sense, but the paper reports no values for them.

free parameters (3)
  • w1-w4 embedding mixture weights = not reported
    Weights in Edoc = w1*Eraw + w2*Eco-occur + w3*ESQL + w4*EsynthQ, learned once per workload by gradient descent on synthetic questions (Section 5.2).
  • Context allocation variables p, ptbl, pcol = not reported
    Bayesian optimization variables determining token split among table, column, and hint documents, fitted on the synthetic workload (Section 6).
  • Bandit epsilon and sliding window length = not reported
    Hyperparameters of the abstention policy (Section 7); values are not specified in the paper.
assumptions (4)
  • domain assumption Past SQL workloads contain reusable subcomponents such as join paths, filter predicates, and group-by clauses that recur in future questions.
    Core premise of the paper, stated in Section 3.2 and Section 4.2, and used to justify query hint documents.
  • domain assumption LLM-generated synthetic questions derived from past SQL queries are representative of real user questions.
    Used in Section 5.1 and Section 6 as the training signal for embedding weights and context allocation. No validation of this assumption is provided.
  • domain assumption Binary user feedback is available at runtime and reflects whether the generated SQL was correct or useful.
    The abstention policy in Section 7 requires a thumbs up or thumbs down signal for every answered question.
  • domain assumption Cosine similarity between embeddings is a valid proxy for whether a document is relevant to a question.
    Used throughout Sections 5 and 6 as the retrieval objective and as the basis for the embedding optimization loss.

how reviews work

0 comments
Cite this review

Pith. "Pith review of TailorSQL: An NL2SQL System Tailored to Your Query Workload." pith.science (2026). https://pith.science/paper/TYU6GJU2

@misc{pith2026250523039,
  author       = {Pith},
  title        = {Pith review of: TailorSQL: An NL2SQL System Tailored to Your Query Workload},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/TYU6GJU2}},
  note         = {Machine review of arXiv:2505.23039}
}
abstract

NL2SQL (natural language to SQL) translates natural language questions into SQL queries, thereby making structured data accessible to non-technical users, serving as the foundation for intelligent data applications. State-of-the-art NL2SQL techniques typically perform translation by retrieving database-specific information, such as the database schema, and invoking a pre-trained large language model (LLM) using the question and retrieved information to generate the SQL query. However, existing NL2SQL techniques miss a key opportunity which is present in real-world settings: NL2SQL is typically applied on existing databases which have already served many SQL queries in the past. The past query workload implicitly contains information which is helpful for accurate NL2SQL translation and is not apparent from the database schema alone, such as common join paths and the semantics of obscurely-named tables and columns. We introduce TailorSQL, a NL2SQL system that takes advantage of information in the past query workload to improve both the accuracy and latency of translating natural language questions into SQL. By specializing to a given workload, TailorSQL achieves up to 2$\times$ improvement in execution accuracy on standardized benchmarks.

Figures

Figures reproduced from arXiv: 2505.23039 by the authors.

Figure 1
Figure 1. (A) The input to NL2SQL is a user question over a database which has a certain schema and past query workload, and the desired output [PITH_FULL_IMAGE:figures/full_fig_p002_1.png] view at source ↗
Figure 2
Figure 2. In an offline process, TailorSQL (A) generates schema and query hint documents based on information in the database, then (B) generates [PITH_FULL_IMAGE:figures/full_fig_p004_2.png] view at source ↗
Figure 3
Figure 3. Past filter predicates help interpret column values. [PITH_FULL_IMAGE:figures/full_fig_p004_3.png] view at source ↗
Figures from the paper (4 more)
Figure 4
Figure 4. Figure 4: Past group-by clauses help indicate an implicit primary key. [PITH_FULL_IMAGE:figures/full_fig_p005_4.png]
Figure 5
Figure 5. Figure 5: TailorSQL achieves 12.5%, 10.9% and 22.7% higher match execution accuracy compared to the next best baseline (SBERT). TailorSQL [PITH_FULL_IMAGE:figures/full_fig_p010_5.png]
Figure 6
Figure 6. Figure 6: TailorSQL’s tailored document embeddings achieve higher [PITH_FULL_IMAGE:figures/full_fig_p011_6.png]
Figure 7
Figure 7. Figure 7: Improvement in recall achieved by including each individual [PITH_FULL_IMAGE:figures/full_fig_p011_7.png]

Discussion (0). Sign in to comment.

Reference graph

Works this paper leans on

40 extracted references · 12 canonical work pages

  1. [1]

    https://learn.microsoft.com/en-us/azure/azure-sql/copilot/query-editor- natural-language-to-sql-copilot

    [n.d.]. https://learn.microsoft.com/en-us/azure/azure-sql/copilot/query-editor- natural-language-to-sql-copilot

    work page

  2. [2]

    https://docs.aws.amazon.com/redshift/latest/mgmt/query-editor-v2- generative-ai.html

    [n.d.]. https://docs.aws.amazon.com/redshift/latest/mgmt/query-editor-v2- generative-ai.html

    work page

  3. [3]

    https://docs.snowflake.com/en/user-guide/snowflake-copilot

    [n.d.]. https://docs.snowflake.com/en/user-guide/snowflake-copilot

    work page

  4. [4]

    https://huggingface.co/sentence-transformers/all-MiniLM-L6-v2

    2022. https://huggingface.co/sentence-transformers/all-MiniLM-L6-v2

    work page 2022

  5. [5]

    Zeynep Akkalyoncu Yilmaz, Wei Yang, Haotian Zhang, and Jimmy Lin. 2019. Cross-Domain Modeling of Sentence-Level Evidence for Document Retrieval. In Proceedings of the 2019 Conference on Empirical Methods in Natural Language Processing and the 9th International Joint Conference on Natural Language Processing (EMNLP-IJCNLP), Kentaro Inui, Jing Jiang, Vincen...

    work page doi:10.18653/v1/d19-1352 2019

  6. [6]

    Jacob Devlin, Ming-Wei Chang, Kenton Lee, and Kristina Toutanova. 2019. BERT: Pre-training of Deep Bidirectional Transformers for Language Understanding. arXiv:1810.04805 [cs.CL]

    arXiv 2019

  7. [7]

    Dawei Gao, Haibin Wang, Yaliang Li, Xiuyu Sun, Yichen Qian, Bolin Ding, and Jingren Zhou. 2023. Text-to-SQL Empowered by Large Language Models: A Benchmark Evaluation. arXiv:2308.15363 [cs.DB]

    arXiv 2023

  8. [8]

    Yingqi Gao, Yifu Liu, Xiaoxia Li, Xiaorong Shi, Yin Zhu, Yiming Wang, Shiqi Li, Wei Li, Yuntao Hong, Zhiling Luo, Jinyang Gao, Liyu Mou, and Yu Li. 2025. A Preview of XiYan-SQL: A Multi-Generator Ensemble Framework for Text-to-SQL. arXiv:2411.08599 [cs.AI] https://arxiv.org/abs/2411.08599

    arXiv 2025

Show all 40 references
  1. [9]

    Zijin Hong, Zheng Yuan, Qinggang Zhang, Hao Chen, Junnan Dong, Feiran Huang, and Xiao Huang. 2025. Next-Generation Database Interfaces: A Survey of LLM- based Text-to-SQL. arXiv:2406.08426 [cs.CL] https://arxiv.org/abs/2406.08426

    2025

  2. [10]

    Anthropic Inc. 2024. https://www.anthropic.com/news/claude-3-haiku

    2024

  3. [11]

    Patrick Lewis, Ethan Perez, Aleksandra Piktus, Fabio Petroni, Vladimir Karpukhin, Naman Goyal, Heinrich Küttler, Mike Lewis, Wen tau Yih, Tim Rocktäschel, Sebastian Riedel, and Douwe Kiela. 2021. Retrieval-Augmented Generation for Knowledge-Intensive NLP Tasks. arXiv:2005.1140...

    2021 arXiv

  4. [12]

    Haoyang Li, Jing Zhang, Hanbing Liu, Ju Fan, Xiaokang Zhang, Jun Zhu, Renjie Wei, Hongyan Pan, Cuiping Li, and Hong Chen. 2024. CodeS: Towards Building Open-source Language Models for Text-to-SQL. Proc. ACM Manag. Data 2, 3, Article 127 (May 2024), 28 pages. https://doi.org/10...

    2024 doi

  5. [13]

    Jinyang Li, Binyuan Hui, Ge Qu, Jiaxi Yang, Binhua Li, Bowen Li, Bailin Wang, Bowen Qin, Rongyu Cao, Ruiying Geng, Nan Huo, Xuanhe Zhou, Chenhao Ma, Guoliang Li, Kevin C. C. Chang, Fei Huang, Reynold Cheng, and Yongbin Li. 2023. Can LLM Already Serve as A Database Interface? A...

    2023 arXiv

  6. [14]

    Nelson F Liu, Kevin Lin, John Hewitt, Ashwin Paranjape, Michele Bevilacqua, Fabio Petroni, and Percy Liang. 2023. Lost in the middle: How language models use long contexts. arXiv preprint arXiv:2307.03172 (2023)

    2023 arXiv

  7. [15]

    Kyle Luoma and Arun Kumar. 2025. SNAILS: Schema Naming Assessments for Improved LLM-Based SQL Inference. Proc. ACM Manag. Data 3, 1, Article 77 (Feb. 2025), 26 pages. https://doi.org/10.1145/3709727

    2025 doi

  8. [16]

    Sean MacAvaney, Andrew Yates, Arman Cohan, and Nazli Goharian. 2019. CEDR: Contextualized Embeddings for Document Ranking. InProceedings of the 42nd International ACM SIGIR Conference on Research and Development in Information Retrieval (SIGIR ’19). ACM. https://doi.org/10.114...

    2019

  9. [17]

    Rodrigo Nogueira and Kyunghyun Cho. 2020. Passage Re-ranking with BERT. arXiv:1901.04085 [cs.IR]

    2020 arXiv

  10. [18]

    Long Ouyang, Jeffrey Wu, Xu Jiang, Diogo Almeida, Carroll Wainwright, Pamela Mishkin, Chong Zhang, Sandhini Agarwal, Katarina Slama, Alex Ray, et al. 2022. Training language models to follow instructions with human feedback.Advances in Neural Information Processing Systems 35 ...

    2022

  11. [19]

    Mohammadreza Pourreza, Hailong Li, Ruoxi Sun, Yeounoh Chung, Shayan Talaei, Gaurav Tarlok Kakkar, Yu Gan, Amin Saberi, Fatma Ozcan, and Sercan O. Arik

  12. [20]

    Mohammadreza Pourreza and Davood Rafiei. 2023. DIN-SQL: Decomposed In- Context Learning of Text-to-SQL with Self-Correction. arXiv:2304.11015 [cs.CL]

    2023 arXiv

  13. [21]

    Rafael Rafailov, Archit Sharma, Eric Mitchell, Stefano Ermon, Christopher D Manning, and Chelsea Finn. 2023. Direct preference optimization: Your language model is secretly a reward model.arXiv preprint arXiv:2305.18290 (2023)

    2023 arXiv

  14. [22]

    Nils Reimers and Iryna Gurevych. 2019. Sentence-BERT: Sentence Embeddings using Siamese BERT-Networks. InConference on Empirical Methods in Natural Language Processing. https://api.semanticscholar.org/CorpusID:201646309

    2019

  15. [23]

    Stephen Robertson and Hugo Zaragoza. 2009. The Probabilistic Relevance Framework: BM25 and Beyond. Found. Trends Inf. Retr.3, 4 (April 2009), 333–389. https://doi.org/10.1561/1500000019

    2009 doi

  16. [24]

    Torsten Scholak, Nathan Schucher, and Dzmitry Bahdanau. 2021. PICARD: Parsing Incrementally for Constrained Auto-Regressive Decoding from Language Models. arXiv:2109.05093 [cs.CL]

    2021 arXiv

  17. [25]

    Jaydeep Sen, Chuan Lei, Abdul Quamar, Fatma Ozcan, Vasilis Efthymiou, Ayushi Dalmia, Greg Stager, Ashish Mittal, Diptikalyan Saha, and Karthik Sankaranarayanan. 2020. ATHENA++: Natural Language Querying for Complex Nested SQL Queries. Proc. VLDB Endow. 13, 11 (2020), 2747–2759

    2020

  18. [26]

    Freda Shi, Xinyun Chen, Kanishka Misra, Nathan Scales, David Dohan, Ed H Chi, Nathanael Schärli, and Denny Zhou. 2023. Large language models can be easily distracted by irrelevant context. InInternational Conference on Machine Learning. PMLR, 31210–31227

    2023

  19. [27]

    Shayan Talaei, Mohammadreza Pourreza, Yu-Chen Chang, Azalia Mirhoseini, and Amin Saberi. 2024. CHESS: Contextual Harnessing for Efficient SQL Synthesis. arXiv:2405.16755 [cs.LG] https://arxiv.org/abs/2405.16755

    2024 arXiv

  20. [28]

    Rohan Taori, Ishaan Gulrajani, Tianyi Zhang, Yann Dubois, Xuechen Li, Carlos Guestrin, Percy Liang, and Tatsunori B Hashimoto. 2023. Alpaca: A strong, replicable instruction-following model.Stanford Center for Research on Foundation Models. https://crfm. stanford. edu/2023/03/...

    2023

  21. [29]

    Hugo Touvron, Louis Martin, Kevin Stone, Peter Albert, Amjad Almahairi, Yasmine Babaei, Nikolay Bashlykov, Soumya Batra, Prajjwal Bhargava, Shruti Bhosale, et al. 2023. Llama 2: Open foundation and fine-tuned chat models.arXiv preprint arXiv:2307.09288 (2023)

    2023 arXiv

  22. [30]

    Alexander van Renen, Dominik Horn, Pascal Pfeil, Kapil Eknath Vaidya, Wenjian Dong, Murali Narayanaswamy, Zhengchun Liu, Gaurav Saxena, Andreas Kipf, and Tim Kraska. 2024. Why TPC is not enough: An analysis of the Amazon Redshift fleet. In VLDB 2024. https://www.amazon.science...

    2024

  23. [31]

    Adrian Vogelsgesang, Michael Haubenschild, Jan Finis, Alfons Kemper, Viktor Leis, Tobias Muehlbauer, Thomas Neumann, and Manuel Then. 2018. Get Real: How Benchmarks Fail to Represent the Real World. In Proceedings of the Workshop on Testing Database Systems (Houston, TX, USA) ...

    2018

  24. [32]

    Bing Wang, Changyu Ren, Jian Yang, Xinnian Liang, Jiaqi Bai, Qian-Wen Zhang, Zhao Yan, and Zhoujun Li. 2023. Mac-sql: Multi-agent collaboration for text-to-sql. arXiv preprint arXiv:2312.11242 (2023)

    2023 arXiv

  25. [33]

    Bailin Wang, Richard Shin, Xiaodong Liu, Oleksandr Polozov, and Matthew Richardson. 2021. RAT-SQL: Relation-Aware Schema Encoding and Linking for Text-to-SQL Parsers. arXiv:1911.04942 [cs.CL]

    2021 arXiv

  26. [34]

    Xuezhi Wang, Jason Wei, Dale Schuurmans, Quoc Le, Ed Chi, Sharan Narang, Aakanksha Chowdhery, and Denny Zhou. 2023. Self-Consistency Improves Chain of Thought Reasoning in Language Models. arXiv:2203.11171 [cs.CL] https://arxiv.org/abs/2203.11171

    2023 arXiv

  27. [35]

    Yizhong Wang, Swaroop Mishra, Pegah Alipoormolabashi, Yeganeh Kordi, Amirreza Mirzaei, Anjana Arunkumar, Arjun Ashok, Arut Selvan Dhanasekaran, Atharva Naik, David Stap, et al. 2022. Super-naturalinstructions: Generalization via declarative instructions on 1600+ nlp tasks.arXi...

    2022 arXiv

  28. [36]

    Jason Wei, Xuezhi Wang, Dale Schuurmans, Maarten Bosma, Brian Ichter, Fei Xia, Ed Chi, Quoc Le, and Denny Zhou. 2023. Chain-of-Thought Prompting Elicits Reasoning in Large Language Models. arXiv:2201.11903 [cs.CL]

    2023 arXiv

  29. [37]

    Tao Yu, Rui Zhang, Kai Yang, Michihiro Yasunaga, Dongxu Wang, Zifan Li, James Ma, Irene Li, Qingning Yao, Shanelle Roman, Zilin Zhang, and Dragomir Radev. 2019. Spider: A Large-Scale Human-Labeled Dataset for Complex and Cross-Domain Semantic Parsing and Text-to-SQL Task. arXi...

    2019 arXiv

  30. [38]

    Chunting Zhou, Pengfei Liu, Puxin Xu, Srini Iyer, Jiao Sun, Yuning Mao, Xuezhe Ma, Avia Efrat, Ping Yu, Lili Yu, et al. 2023. Lima: Less is more for alignment. arXiv preprint arXiv:2305.11206 (2023)

    2023 arXiv

  31. [39]

    Denny Zhou, Nathanael Schärli, Le Hou, Jason Wei, Nathan Scales, Xuezhi Wang, Dale Schuurmans, Claire Cui, Olivier Bousquet, Quoc Le, and Ed Chi. 2023. Least-to-Most Prompting Enables Complex Reasoning in Large Language Models. arXiv:2205.10625 [cs.AI]

    2023 arXiv

  32. [2024]

    arXiv:2410.01943 [cs.LG] https://arxiv.org/abs/2410.01943

    CHASE-SQL: Multi-Path Reasoning and Preference Optimized Candidate Se- lection in Text-to-SQL. arXiv:2410.01943 [cs.LG] https://arxiv.org/abs/2410.01943

    arXiv

Pith tools

Reviewed August 7, 2026 · model on record in the stance chip above.