Are UFOs Driving Innovation? The Illusion of Causality in Large Language Models

TL;DR

This study investigates whether large language models develop illusions of causality, finding that some models are more susceptible to this bias, especially when influenced by sycophantic prompts, with Claude-3.5-Sonnet being the most robust.

Contribution

The paper introduces a novel evaluation of causality illusions in large language models and compares their susceptibility, highlighting model differences and the impact of prompt manipulation.

Findings

Claude-3.5-Sonnet shows the lowest causal illusion susceptibility.

Sycophantic prompts increase causal illusions, especially in GPT-4o-Mini.

Claude-3.5-Sonnet remains most robust against causal bias.

Abstract

Illusions of causality occur when people develop the belief that there is a causal connection between two variables with no supporting evidence. This cognitive bias has been proposed to underlie many societal problems including social prejudice, stereotype formation, misinformation and superstitious thinking. In this research we investigate whether large language models develop the illusion of causality in real-world settings. We evaluated and compared news headlines generated by GPT-4o-Mini, Claude-3.5-Sonnet, and Gemini-1.5-Pro to determine whether the models incorrectly framed correlations as causal relationships. In order to also measure sycophantic behavior, which occurs when a model aligns with a user's beliefs in order to look favorable even if it is not objectively correct, we additionally incorporated the bias into the prompts, observing if this manipulation increases the likelihood of the models exhibiting the illusion of causality. We found that Claude-3.5-Sonnet is the model that presents the lowest degree of causal illusion aligned with experiments on Correlation-to-Causation Exaggeration in human-written press releases. On the other hand, our findings suggest that while mimicry sycophancy increases the likelihood of causal illusions in these models, especially in GPT-4o-Mini, Claude-3.5-Sonnet remains the most robust against this cognitive bias.