Primacy Effect of ChatGPT

October 20, 2023 · Entered Twilight · 🏛 Conference on Empirical Methods in Natural Language Processing

Repo contents: .gitattributes, Dataset, README.md, experiment.ipynb, fig_1.png, fig_2.png, fig_3.png

Authors Yiwei Wang, Yujun Cai, Muhao Chen, Yuxuan Liang, Bryan Hooi arXiv ID 2310.13206 Category cs.CL: Computation & Language Cross-listed cs.AI Citations 27 Venue Conference on Empirical Methods in Natural Language Processing Repository https://github.com/wangywUST/PrimacyEffectGPT ⭐ 4 Last Checked 2 months ago

Abstract

Instruction-tuned large language models (LLMs), such as ChatGPT, have led to promising zero-shot performance in discriminative natural language understanding (NLU) tasks. This involves querying the LLM using a prompt containing the question, and the candidate labels to choose from. The question-answering capabilities of ChatGPT arise from its pre-training on large amounts of human-written text, as well as its subsequent fine-tuning on human preferences, which motivates us to ask: Does ChatGPT also inherits humans' cognitive biases? In this paper, we study the primacy effect of ChatGPT: the tendency of selecting the labels at earlier positions as the answer. We have two main findings: i) ChatGPT's decision is sensitive to the order of labels in the prompt; ii) ChatGPT has a clearly higher chance to select the labels at earlier positions as the answer. We hope that our experiments and analyses provide additional insights into building more reliable ChatGPT-based solutions. We release the source code at https://github.com/wangywUST/PrimacyEffectGPT.