How Far Can We Extract Diverse Perspectives from Large Language Models?
This addresses the challenge of costly human opinion collection by exploring scalable LLM-based solutions for generating diverse data, though it is incremental in improving prompting techniques.
The study investigated the capacity of large language models (LLMs) to generate diverse perspectives on subjective topics like social norms, finding that LLMs can produce diverse opinions based on task subjectivity and perform on par with humans in extracting maximum diversity.
Collecting diverse human opinions is costly and challenging. This leads to a recent trend in exploiting large language models (LLMs) for generating diverse data for potential scalable and efficient solutions. However, the extent to which LLMs can generate diverse perspectives on subjective topics is still unclear. In this study, we explore LLMs' capacity of generating diverse perspectives and rationales on subjective topics such as social norms and argumentative texts. We introduce the problem of extracting maximum diversity from LLMs. Motivated by how humans form opinions based on values, we propose a criteria-based prompting technique to ground diverse opinions. To see how far we can extract diverse perspectives from LLMs, or called diversity coverage, we employ a step-by-step recall prompting to generate more outputs from the model iteratively. Our methods, applied to various tasks, show that LLMs can indeed produce diverse opinions according to the degree of task subjectivity. We also find that LLM's performance of extracting maximum diversity is on par with human.