Table as a Modality for Large Language Models

TL;DR

This paper introduces TAMO, a multimodal framework that treats tables as an independent modality to improve large language models' ability to reason with tabular data, significantly enhancing performance.

Contribution

The paper proposes TAMO, a novel multimodal approach that integrates tables as a separate modality using a hypergraph neural network with LLMs, addressing structural information loss.

Findings

Achieved an average relative gain of 42.65% across benchmarks.

Significant improvements in generalization on tabular reasoning tasks.

Demonstrated the effectiveness of treating tables as a distinct modality.

Abstract

To migrate the remarkable successes of Large Language Models (LLMs), the community has made numerous efforts to generalize them to the table reasoning tasks for the widely deployed tabular data. Despite that, in this work, by showing a probing experiment on our proposed StructQA benchmark, we postulate that even the most advanced LLMs (such as GPTs) may still fall short of coping with tabular data. More specifically, the current scheme often simply relies on serializing the tabular data, together with the meta information, then inputting them through the LLMs. We argue that the loss of structural information is the root of this shortcoming. In this work, we further propose TAMO, which bears an ideology to treat the tables as an independent modality integrated with the text tokens. The resulting model in TAMO is a multimodal framework consisting of a hypergraph neural network as the global table encoder seamlessly integrated with the mainstream LLM. Empirical results on various benchmarking datasets, including HiTab, WikiTQ, WikiSQL, FeTaQA, and StructQA, have demonstrated significant improvements on generalization with an average relative gain of 42.65%.