RuozhiBench: Evaluating LLMs with Logical Fallacies and Misleading Premises

TL;DR

This paper introduces RuozhiBench, a bilingual dataset designed to evaluate large language models' ability to identify and reason about logical fallacies and misleading premises, revealing significant gaps in current models' reasoning skills.

Contribution

The paper presents RuozhiBench, a novel challenging dataset for assessing LLMs' reasoning about deceptive reasoning, and provides a comprehensive evaluation of 17 models' performance on this task.

Findings

LLMs perform poorly on detecting logical fallacies, with the best model achieving only 62% accuracy.

Models show high scores on traditional benchmarks but struggle with deceptive reasoning.

There is a significant gap between LLMs and human performance in reasoning about logical fallacies.

Abstract

Recent advances in large language models (LLMs) have shown that they can answer questions requiring complex reasoning. However, their ability to identify and respond to text containing logical fallacies or deliberately misleading premises remains less studied. To address this gap, we introduce RuozhiBench, a bilingual dataset comprising 677 carefully curated questions that contain various forms of deceptive reasoning, meticulously crafted through extensive human effort and expert review. In a comprehensive evaluation of 17 LLMs from 5 Series over RuozhiBench using both open-ended and two-choice formats, we conduct extensive analyses on evaluation protocols and result patterns. Despite their high scores on conventional benchmarks, these models showed limited ability to detect and reason correctly about logical fallacies, with even the best-performing model, Claude-3-haiku, achieving only 62% accuracy compared to the human of more than 90%.