Exploring the Numerical Reasoning Capabilities of Language Models: A Comprehensive Analysis on Tabular Data

TL;DR

This paper provides a detailed taxonomy and evaluation of numerical reasoning skills in language models, revealing their strengths and weaknesses across various reasoning types, especially in tabular data tasks.

Contribution

It introduces a hierarchical taxonomy of numerical reasoning skills, develops diverse probes, and evaluates state-of-the-art models on tabular inference, highlighting their limitations and strengths.

Findings

No model excels across all reasoning types.

FlanT5 and GPT-3.5 show strong overall reasoning skills.

Models often exploit dataset artifacts for predictions.

Abstract

Numbers are crucial for various real-world domains such as finance, economics, and science. Thus, understanding and reasoning with numbers are essential skills for language models to solve different tasks. While different numerical benchmarks have been introduced in recent years, they are limited to specific numerical aspects mostly. In this paper, we propose a hierarchical taxonomy for numerical reasoning skills with more than ten reasoning types across four levels: representation, number sense, manipulation, and complex reasoning. We conduct a comprehensive evaluation of state-of-the-art models to identify reasoning challenges specific to them. Henceforth, we develop a diverse set of numerical probes employing a semi-automated approach. We focus on the tabular Natural Language Inference (TNLI) task as a case study and measure models' performance shifts. Our results show that no model consistently excels across all numerical reasoning types. Among the probed models, FlanT5 (few-/zero-shot) and GPT-3.5 (few-shot) demonstrate strong overall numerical reasoning skills compared to other models. Label-flipping probes indicate that models often exploit dataset artifacts to predict the correct labels.