Share your thoughts, 1 month free Claude Pro on usSee more
WorkDL logo mark

TabLLM: Few-shot Classification of Tabular Data with Large Language Models

About

We study the application of large language models to zero-shot and few-shot classification of tabular data. We prompt the large language model with a serialization of the tabular data to a natural-language string, together with a short description of the classification problem. In the few-shot setting, we fine-tune the large language model using some labeled examples. We evaluate several serialization methods including templates, table-to-text models, and large language models. Despite its simplicity, we find that this technique outperforms prior deep-learning-based tabular classification methods on several benchmark datasets. In most cases, even zero-shot classification obtains non-trivial performance, illustrating the method's ability to exploit prior knowledge encoded in large language models. Unlike many deep learning methods for tabular datasets, this approach is also competitive with strong traditional baselines like gradient-boosted trees, especially in the very-few-shot setting.

Stefan Hegselmann, Alejandro Buendia, Hunter Lang, Monica Agrawal, Xiaoyi Jiang, David Sontag• 2022

Related benchmarks

TaskDatasetResultRank
Tabular Classification75 Tabular Classification Datasets (test)
Accuracy41.6
89
Tabular Regression52 Tabular Datasets (test)
NMAE0.476
85
ClassificationHI
Accuracy0.54
59
ClassificationCredit-g
ROC AUC0.6038
53
Classificationblood
ROC-AUC0.6601
47
Aggregate Tabular BenchmarkingAggregate
Avg Rank13.31
33
Tabular ClassificationHeart
Mean AUC-ROC81.72
31
Tabular ClassificationAdult
AUC0.8357
24
Tabular ClassificationAMAZON
AUC0.5162
24
Tabular ClassificationDiabetes
AUC70.42
24
Showing 10 of 20 rows

Other info

Follow for update