Large Language Models are few(1)-shot Table Reasoners

TL;DR

This paper investigates the ability of large language models to perform table reasoning tasks with few-shot learning, demonstrating their strong performance and reasoning capabilities even without training on table data.

Contribution

It shows that LLMs can effectively perform complex table reasoning with minimal examples and chain-of-thought prompting, serving as a strong baseline for future research.

Findings

LLMs perform well on table QA and fact verification datasets.

Chain of thought prompting significantly improves LLM performance.

Reasoning chains from LLMs align well with semantic structures.

Abstract

Recent literature has shown that large language models (LLMs) are generally excellent few-shot reasoners to solve text reasoning tasks. However, the capability of LLMs on table reasoning tasks is yet to be explored. In this paper, we aim at understanding how well LLMs can perform table-related tasks with few-shot in-context learning. Specifically, we evaluated LLMs on popular table QA and fact verification datasets like WikiTableQuestion, FetaQA, TabFact, and FEVEROUS and found that LLMs are competent at complex reasoning over table structures, though these models are not pre-trained on any table corpus. When combined with `chain of thoughts' prompting, LLMs can achieve very strong performance with only a 1-shot demonstration, even on par with some SoTA models. We show that LLMs are even more competent at generating comprehensive long-form answers on FetaQA than tuned T5-large. We further manually studied the reasoning chains elicited from LLMs and found that these reasoning chains are highly consistent with the underlying semantic form. We believe that LLMs can serve as a simple yet generic baseline for future research. The code and data are released in https://github.com/wenhuchen/TableCoT.