Cognitive Effects in Large Language Models

TL;DR

This study investigates whether GPT-3, a large language model, exhibits human-like cognitive biases and effects, revealing that it mimics several but not all human cognitive patterns.

Contribution

The paper introduces a methodology for testing cognitive effects in LLMs and demonstrates that GPT-3 exhibits multiple human cognitive biases, highlighting similarities and differences with human cognition.

Findings

GPT-3 shows priming, distance, SNARC, and size congruity effects.

GPT-3 does not exhibit the anchoring effect.

Methodology for converting real-world experiments to text-based experiments.

Abstract

Large Language Models (LLMs) such as ChatGPT have received enormous attention over the past year and are now used by hundreds of millions of people every day. The rapid adoption of this technology naturally raises questions about the possible biases such models might exhibit. In this work, we tested one of these models (GPT-3) on a range of cognitive effects, which are systematic patterns that are usually found in human cognitive tasks. We found that LLMs are indeed prone to several human cognitive effects. Specifically, we show that the priming, distance, SNARC, and size congruity effects were presented with GPT-3, while the anchoring effect is absent. We describe our methodology, and specifically the way we converted real-world experiments to text-based experiments. Finally, we speculate on the possible reasons why GPT-3 exhibits these effects and discuss whether they are imitated or reinvented.