Back to papers
Weak-to-Strong Prompts with Lightweight-to-Powerful LLMs for High-Accuracy, Low-Cost, and Explainable Data Transformation
Summary: MegaTran uses a fine-tuned lightweight LLM to turn loose user prompts into structured task descriptions (Weak2StrongPrompt), then a powerful LLM generates transformation code (Prompt2Code). Sanity-check reflection with checklists and LazyRAG retrieval of external snippets make results explainable, low-cost, and 2.2–26.1% more accurate than prior methods.
(summarized by gpt-5-mini on Feb 09 2026)
- Paper ID
- 13886
- Venue
- VLDB
- Year
- 2025
- Pagerank
- 4.1905499e-05
- Overall Rank
- 10,618 | 26.21%
- DOI
-
10.14778/3742728.3742734
Incoming Non-self Citations Over Time
No non-self incoming citations found for this paper in this database.
Incoming Citations (Sorted by Pagerank)
Showing 1 of 1 citing papers.
Outgoing Citations (Sorted by Pagerank)
Showing 20 of 20 cited papers.
Citations counted here include only citations to other VLDB/SIGMOD/CIDR/PODS papers in this database.
| Rank |
Cited Paper |
Year |
Venue |
Pagerank |
| 112 |
Potter's Wheel: An Interactive Data Cleaning System |
2001 |
VLDB |
0.00047010908 |
| 516 |
Can Foundation Models Wrangle Your Data? |
2023 |
VLDB |
0.00021194444 |
| 1,012 |
NADEEF: A Commodity Data Cleaning System |
2013 |
SIGMOD |
0.00014638349 |
| 1,266 |
Foofah: Transforming Data By Example |
2017 |
SIGMOD |
0.00012921038 |
| 1,274 |
The Data Civilizer System |
2017 |
CIDR |
0.00012869297 |
| 1,469 |
BlinkFill: Semi-supervised Programming By Example for Syntactic String Transformations |
2016 |
VLDB |
0.00011831156 |
| 1,821 |
Synthesizing Entity Matching Rules by Examples |
2018 |
VLDB |
0.00010406856 |
| 2,348 |
RPT: Relational Pre-trained Transformer Is Almost All You Need towards Democratizing Data Preparation |
2021 |
VLDB |
8.9903659e-05 |
| 3,003 |
Chorus: Foundation Models for Unified Data Discovery and Exploration |
2024 |
VLDB |
7.7358219e-05 |
| 3,198 |
Towards Dependable Data Repairing with Fixing Rules |
2014 |
SIGMOD |
7.4029546e-05 |
| 3,483 |
Transform-Data-by-Example (TDE): An Extensible Search Engine for Data Transformations |
2018 |
VLDB |
7.0493668e-05 |
| 4,211 |
Unicorn: A Unified Multi-tasking Model for Supporting Matching Tasks in Data Integration |
2023 |
SIGMOD |
6.3495931e-05 |
| 4,908 |
Combining Small Language Models and Large Language Models for Zero-Shot NL2SQL |
2024 |
VLDB |
5.835596e-05 |
| 5,096 |
Auto-Transform: Learning-to-Transform by Patterns |
2020 |
VLDB |
5.6960764e-05 |
| 5,283 |
Explaining Dataset Changes for Semantic Data Versioning with Explain-Da-V |
2023 |
VLDB |
5.5843052e-05 |
| 5,984 |
DataPrep.EDA: Task-Centric Exploratory Data Analysis for Statistical Modeling in Python |
2021 |
SIGMOD |
5.2400405e-05 |
| 6,798 |
DTT: An Example-Driven Tabular Transformer for Joinability by Leveraging Large Language Models |
2024 |
SIGMOD |
4.9186164e-05 |
| 7,458 |
Trinity: An Extensible Synthesis Framework for Data Science |
2019 |
VLDB |
4.7200076e-05 |
| 7,992 |
Data Civilizer 2.0: A Holistic Framework for Data Preparation and Analytics |
2019 |
VLDB |
4.6082565e-05 |
| 9,576 |
CoClean: Collaborative Data Cleaning |
2020 |
SIGMOD |
4.3207e-05 |
Semantically Similar Papers
| Overall Rank |
Paper |
Year |
Venue |
Pagerank |
| 10,603 |
Optimized Batch Prompting for Cost-effective LLMs |
2025 |
VLDB |
4.1905499e-05 |
| 6,798 |
DTT: An Example-Driven Tabular Transformer for Joinability by Leveraging Large Language Models |
2024 |
SIGMOD |
4.9186164e-05 |
| 6,202 |
Chat2Data: An Interactive Data Analysis System with RAG, Vector Databases and LLMs |
2024 |
VLDB |
5.1554849e-05 |
| 8,157 |
Automated Data Visualization from Natural Language via Large Language Models: An Exploratory Study |
2024 |
SIGMOD |
4.5701714e-05 |
| 10,064 |
Cut Costs, Not Accuracy: LLM-Powered Data Processing with Guarantees |
2026 |
SIGMOD |
4.1905499e-05 |
| 1,088 |
Language Models Enable Simple Systems for Generating Structured Views of Heterogeneous Data Lakes |
2024 |
VLDB |
0.00014158762 |
| 10,328 |
LLM-AutoDP: Automatic Data Processing via LLM Agents for Model Fine-tuning |
2026 |
VLDB |
4.1905499e-05 |
| 3,803 |
Revisiting Prompt Engineering via Declarative Crowdsourcing |
2024 |
CIDR |
6.7498941e-05 |
| 9,405 |
TabulaX: Leveraging Large Language Models for Multi-Class Table Transformations |
2025 |
VLDB |
4.3399748e-05 |
| 7,016 |
LLM for Data Management |
2024 |
VLDB |
4.8561622e-05 |