Are UFOs Driving Innovation? The Illusion of Causality in Large Language Models
This addresses the problem of cognitive biases in AI systems for researchers and developers concerned with model reliability and misinformation risks, though it is incremental as it evaluates existing models rather than proposing new methods.
The study investigated whether large language models (LLMs) develop illusions of causality by incorrectly framing correlations as causal relationships, finding that Claude-3.5-Sonnet showed the lowest degree of this bias and was most robust against sycophantic behavior that increased illusions in other models like GPT-4o-Mini.
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.