@inproceedings{yan-etal-2026-bottleneck,
title = "When the Bottleneck Shifts: Diagnosing and Closing Information Gaps in Interactive Text-to-{SQL}",
author = "Yan, Mengyi and
Ma, Zhinian and
Wang, Yue and
Wang, Yaoshu and
Zhu, Xiaoke and
Wang, Zizhe and
Zhou, Haoyi and
Chen, Zhumin",
editor = "Lee, Hung-Yi and
Dev, Sunipa and
Li, Jessy and
Elliott, Desmond",
booktitle = "Proceedings of the 2026 Conference on Empirical Methods in Natural Language Processing",
month = oct,
year = "2026",
address = "Budapest, Hungary",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2026.emnlp-main.465/",
pages = "10655--10687",
ISBN = "979-8-89176-459-0",
abstract = "In interactive text-to-SQL, the evidence needed to write correct SQL is not given upfront but distributed across schema, knowledge bases, and users, and must be acquired through costly environment interactions under a finite budget. We observe that strong LLMs often fail not from deficient SQL synthesis, but from premature commitment to SQL before acquiring task-critical evidence. Through controlled trajectory analysis, we formalize this observation as the Information Sufficiency Hypothesis and identify a five-class gap taxonomy that structures the acquisition space. Guided by this diagnosis, we develop PASCAL, a training-free acquisition-control data agent framework that interposes a source-aware information-acquisition layer before SQL generation and schedules acquisition by cost-normalized priority. On BIRD-Interact, PASCAL with a local 35B backbone achieves 51.0{\%} Phase-1 and 37.0{\%} two-phase success rate, outperforming frontier models leaderboard result by up to 14.7 percentage points, with gains concentrated on BI tasks requiring cross-source evidence acquisition."
}<?xml version="1.0" encoding="UTF-8"?>
<modsCollection xmlns="http://www.loc.gov/mods/v3">
<mods ID="yan-etal-2026-bottleneck">
<titleInfo>
<title>When the Bottleneck Shifts: Diagnosing and Closing Information Gaps in Interactive Text-to-SQL</title>
</titleInfo>
<name type="personal">
<namePart type="given">Mengyi</namePart>
<namePart type="family">Yan</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Zhinian</namePart>
<namePart type="family">Ma</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Yue</namePart>
<namePart type="family">Wang</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Yaoshu</namePart>
<namePart type="family">Wang</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Xiaoke</namePart>
<namePart type="family">Zhu</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Zizhe</namePart>
<namePart type="family">Wang</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Haoyi</namePart>
<namePart type="family">Zhou</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Zhumin</namePart>
<namePart type="family">Chen</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<originInfo>
<dateIssued>2026-10</dateIssued>
</originInfo>
<typeOfResource>text</typeOfResource>
<relatedItem type="host">
<titleInfo>
<title>Proceedings of the 2026 Conference on Empirical Methods in Natural Language Processing</title>
</titleInfo>
<name type="personal">
<namePart type="given">Hung-Yi</namePart>
<namePart type="family">Lee</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Sunipa</namePart>
<namePart type="family">Dev</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Jessy</namePart>
<namePart type="family">Li</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Desmond</namePart>
<namePart type="family">Elliott</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<originInfo>
<publisher>Association for Computational Linguistics</publisher>
<place>
<placeTerm type="text">Budapest, Hungary</placeTerm>
</place>
</originInfo>
<genre authority="marcgt">conference publication</genre>
<identifier type="isbn">979-8-89176-459-0</identifier>
</relatedItem>
<abstract>In interactive text-to-SQL, the evidence needed to write correct SQL is not given upfront but distributed across schema, knowledge bases, and users, and must be acquired through costly environment interactions under a finite budget. We observe that strong LLMs often fail not from deficient SQL synthesis, but from premature commitment to SQL before acquiring task-critical evidence. Through controlled trajectory analysis, we formalize this observation as the Information Sufficiency Hypothesis and identify a five-class gap taxonomy that structures the acquisition space. Guided by this diagnosis, we develop PASCAL, a training-free acquisition-control data agent framework that interposes a source-aware information-acquisition layer before SQL generation and schedules acquisition by cost-normalized priority. On BIRD-Interact, PASCAL with a local 35B backbone achieves 51.0% Phase-1 and 37.0% two-phase success rate, outperforming frontier models leaderboard result by up to 14.7 percentage points, with gains concentrated on BI tasks requiring cross-source evidence acquisition.</abstract>
<identifier type="citekey">yan-etal-2026-bottleneck</identifier>
<location>
<url>https://aclanthology.org/2026.emnlp-main.465/</url>
</location>
<part>
<date>2026-10</date>
<extent unit="page">
<start>10655</start>
<end>10687</end>
</extent>
</part>
</mods>
</modsCollection>
%0 Conference Proceedings
%T When the Bottleneck Shifts: Diagnosing and Closing Information Gaps in Interactive Text-to-SQL
%A Yan, Mengyi
%A Ma, Zhinian
%A Wang, Yue
%A Wang, Yaoshu
%A Zhu, Xiaoke
%A Wang, Zizhe
%A Zhou, Haoyi
%A Chen, Zhumin
%Y Lee, Hung-Yi
%Y Dev, Sunipa
%Y Li, Jessy
%Y Elliott, Desmond
%S Proceedings of the 2026 Conference on Empirical Methods in Natural Language Processing
%D 2026
%8 October
%I Association for Computational Linguistics
%C Budapest, Hungary
%@ 979-8-89176-459-0
%F yan-etal-2026-bottleneck
%X In interactive text-to-SQL, the evidence needed to write correct SQL is not given upfront but distributed across schema, knowledge bases, and users, and must be acquired through costly environment interactions under a finite budget. We observe that strong LLMs often fail not from deficient SQL synthesis, but from premature commitment to SQL before acquiring task-critical evidence. Through controlled trajectory analysis, we formalize this observation as the Information Sufficiency Hypothesis and identify a five-class gap taxonomy that structures the acquisition space. Guided by this diagnosis, we develop PASCAL, a training-free acquisition-control data agent framework that interposes a source-aware information-acquisition layer before SQL generation and schedules acquisition by cost-normalized priority. On BIRD-Interact, PASCAL with a local 35B backbone achieves 51.0% Phase-1 and 37.0% two-phase success rate, outperforming frontier models leaderboard result by up to 14.7 percentage points, with gains concentrated on BI tasks requiring cross-source evidence acquisition.
%U https://aclanthology.org/2026.emnlp-main.465/
%P 10655-10687
Markdown (Informal)
[When the Bottleneck Shifts: Diagnosing and Closing Information Gaps in Interactive Text-to-SQL](https://aclanthology.org/2026.emnlp-main.465/) (Yan et al., EMNLP 2026)
ACL
- Mengyi Yan, Zhinian Ma, Yue Wang, Yaoshu Wang, Xiaoke Zhu, Zizhe Wang, Haoyi Zhou, and Zhumin Chen. 2026. When the Bottleneck Shifts: Diagnosing and Closing Information Gaps in Interactive Text-to-SQL. In Proceedings of the 2026 Conference on Empirical Methods in Natural Language Processing, pages 10655–10687, Budapest, Hungary. Association for Computational Linguistics.