dorsal/arxiv
View SchemaMemo-SQL: Structured Decomposition and Experience-Driven Self-Correction for Training-Free NL2SQL
| Authors | Zerui Yang, Weichuan Wang, Yanwei Xu, Linqi Song, Yudai Matsuda, Wei Han, Bo Bai |
|---|---|
| Categories | |
| ArXiv ID | 2601.10011vv1 |
| URL | https://arxiv.org/abs/2601.10011 |
| License | http://creativecommons.org/licenses/by/4.0/ |
Abstract
Existing NL2SQL systems face two critical limitations: (1) they rely on in-context learning with only correct examples, overlooking the rich signal in historical error-fix pairs that could guide more robust self-correction; and (2) test-time scaling approaches often decompose questions arbitrarily, producing near-identical SQL candidates across runs and diminishing ensemble gains. Moreover, these methods suffer from a stark accuracy-efficiency trade-off: high performance demands excessive computation, while fast variants compromise quality. We present Memo-SQL, a training-free framework that addresses these issues through two simple ideas: structured decomposition and experience-aware self-correction. Instead of leaving decomposition to chance, we apply three clear strategies, entity-wise, hierarchical, and atomic sequential, to encourage diverse reasoning. For correction, we build a dynamic memory of both successful queries and historical error-fix pairs, and use retrieval-augmented prompting to bring relevant examples into context at inference time, no fine-tuning or external APIs required. On BIRD, Memo-SQL achieves 68.5% execution accuracy, setting a new state of the art among open, zero-fine-tuning methods, while using over 10 times fewer resources than prior TTS approaches.
{
"annotation_id": "28b572fb-3ac6-4a2b-b2c2-180297c8ecc4",
"date_created": "2026-02-17T05:53:23.545000Z",
"date_modified": "2026-02-17T05:53:23.545000Z",
"file_hash": "e7931068624bc55805b66b88ecd8bac89ad1fd998f10905ff6c0b37c6bab3b36",
"private": false,
"record": {
"abstract": "Existing NL2SQL systems face two critical limitations: (1) they rely on in-context learning with only correct examples, overlooking the rich signal in historical error-fix pairs that could guide more robust self-correction; and (2) test-time scaling approaches often decompose questions arbitrarily, producing near-identical SQL candidates across runs and diminishing ensemble gains. Moreover, these methods suffer from a stark accuracy-efficiency trade-off: high performance demands excessive computation, while fast variants compromise quality. We present Memo-SQL, a training-free framework that addresses these issues through two simple ideas: structured decomposition and experience-aware self-correction. Instead of leaving decomposition to chance, we apply three clear strategies, entity-wise, hierarchical, and atomic sequential, to encourage diverse reasoning. For correction, we build a dynamic memory of both successful queries and historical error-fix pairs, and use retrieval-augmented prompting to bring relevant examples into context at inference time, no fine-tuning or external APIs required. On BIRD, Memo-SQL achieves 68.5% execution accuracy, setting a new state of the art among open, zero-fine-tuning methods, while using over 10 times fewer resources than prior TTS approaches.",
"arxiv_id": "2601.10011",
"authors": [
"Zerui Yang",
"Weichuan Wang",
"Yanwei Xu",
"Linqi Song",
"Yudai Matsuda",
"Wei Han",
"Bo Bai"
],
"categories": [
"cs.AI"
],
"license": "http://creativecommons.org/licenses/by/4.0/",
"title": "Memo-SQL: Structured Decomposition and Experience-Driven Self-Correction for Training-Free NL2SQL",
"url": "https://arxiv.org/abs/2601.10011",
"version": "v1"
},
"schema_id": "dorsal/arxiv",
"source": {
"execution_id": "81a27b85-d916-4194-9bbc-bd13096ab39c",
"id": "arXiv Dataset",
"type": "Model",
"variant": "snapshot-2026-01-17",
"version": "0.1.0"
},
"user_id": 1000002
}