Can One-sided Arguments Lead to Response Change in Large Language Models?
This addresses the problem of bias and manipulation in LLM responses for users and developers, but it is incremental as it builds on existing research about LLM behavior.
The study investigated whether large language models (LLMs) can be steered to adopt a specific viewpoint by providing only one-sided arguments, finding that opinion steering occurs across diverse models, argument numbers, and topics, with switching arguments consistently decreasing this effect.
Polemic questions need more than one viewpoint to express a balanced answer. Large Language Models (LLMs) can provide a balanced answer, but also take a single aligned viewpoint or refuse to answer. In this paper, we study if such initial responses can be steered to a specific viewpoint in a simple and intuitive way: by only providing one-sided arguments supporting the viewpoint. Our systematic study has three dimensions: (i) which stance is induced in the LLM response, (ii) how the polemic question is formulated, (iii) how the arguments are shown. We construct a small dataset and remarkably find that opinion steering occurs across (i)-(iii) for diverse models, number of arguments, and topics. Switching to other arguments consistently decreases opinion steering.