Can Models Help Us Create Better Models? Evaluating LLMs as Data Scientists
This provides a new evaluation method for LLMs in data science applications, though it appears incremental relative to existing benchmarking approaches.
The authors tackled the problem of evaluating large language models' ability to write feature engineering code for data science tasks, and demonstrated that their proposed benchmark (FeatEng) can efficiently assess LLM capabilities by measuring XGBoost performance improvements on modified datasets.
We present a benchmark for large language models designed to tackle one of the most knowledge-intensive tasks in data science: writing feature engineering code, which requires domain knowledge in addition to a deep understanding of the underlying problem and data structure. The model is provided with a dataset description in a prompt and asked to generate code transforming it. The evaluation score is derived from the improvement achieved by an XGBoost model fit on the modified dataset compared to the original data. By an extensive evaluation of state-of-the-art models and comparison to well-established benchmarks, we demonstrate that the FeatEng of our proposal can cheaply and efficiently assess the broad capabilities of LLMs, in contrast to the existing methods.