Solving Data-Centric Tasks with Large Language Models

Anonymous

Abstract:

Large language models (LLMs) are rapidly replacing help forums like StackOverflow, and are especially helpful for non-professional programmers and end users. These users are often interested in data-centric tasks, such as spreadsheet manipulation and data wrangling, which are hard to solve if the intent is only communicated using a natural-language description, without including the data. On the other hand, sending large amounts of data to the LLM is intractable and expensive. But how do we decide how much data and which data to include in the prompt?
This paper makes two contributions towards answering this question. First, we create a dataset of real-world NL-to-code tasks manipulating tabular data, with input data tables mined from StackOverflow posts. Second, we introduce a cluster-then-select prompting technique, which adds the most representative rows from the input data to the LLM prompt. We find that the performance and quality of code generated by LLMs is affected by the choice and position of the input data. Our experiments show that LLM performance is indeed sensitive to the amount of data passed in the prompt, and that for tasks with a lot of syntactic variation in the input table, our cluster-then-select technique outperforms a random selection baseline.

Paper Type: short

Research Area: Resources and Evaluation

Contribution Types: Model analysis & interpretability, Data resources, Data analysis

Languages Studied: Python (specifically Pandas)