How susceptible are LLMs to Logical Fallacies?

TL;DR

This paper introduces LOGICOM, a benchmark to evaluate LLMs' robustness against logical fallacies in debates, revealing that GPT-3.5 and GPT-4 are often misled by fallacious arguments.

Contribution

The paper presents LOGICOM, a novel diagnostic benchmark for assessing LLMs' susceptibility to logical fallacies in argumentative debates.

Findings

GPT-3.5 and GPT-4 can change opinions through reasoning

Both models are misled by fallacies 41% and 69% more often

A new dataset of 5,000 logical vs. fallacious argument pairs is provided

Abstract

This paper investigates the rational thinking capability of Large Language Models (LLMs) in multi-round argumentative debates by exploring the impact of fallacious arguments on their logical reasoning performance. More specifically, we present Logic Competence Measurement Benchmark (LOGICOM), a diagnostic benchmark to assess the robustness of LLMs against logical fallacies. LOGICOM involves two agents: a persuader and a debater engaging in a multi-round debate on a controversial topic, where the persuader tries to convince the debater of the correctness of its claim. First, LOGICOM assesses the potential of LLMs to change their opinions through reasoning. Then, it evaluates the debater's performance in logical reasoning by contrasting the scenario where the persuader employs logical fallacies against one where logical reasoning is used. We use this benchmark to evaluate the performance of GPT-3.5 and GPT-4 using a dataset containing controversial topics, claims, and reasons supporting them. Our findings indicate that both GPT-3.5 and GPT-4 can adjust their opinion through reasoning. However, when presented with logical fallacies, GPT-3.5 and GPT-4 are erroneously convinced 41% and 69% more often, respectively, compared to when logical reasoning is used. Finally, we introduce a new dataset containing over 5k pairs of logical vs. fallacious arguments. The source code and dataset of this work are made publicly available.