On the steerability of large language models toward data-driven personas

TL;DR

This paper introduces a data-driven approach to steer large language models toward specific, nuanced personas based on collaborative filtering, improving controllability and diversity of generated viewpoints.

Contribution

It proposes a novel data-driven persona concept and an efficient steering method, significantly enhancing LLMs' ability to generate diverse perspectives.

Findings

Achieved 57-77% improvement in steerability over baselines.

Enabled generation of multiple viewpoints reflecting diverse social groups.

Enhanced control over LLM responses beyond demographic stereotypes.

Abstract

Large language models (LLMs) are known to generate biased responses where the opinions of certain groups and populations are underrepresented. Here, we present a novel approach to achieve controllable generation of specific viewpoints using LLMs, that can be leveraged to produce multiple perspectives and to reflect the diverse opinions. Moving beyond the traditional reliance on demographics like age, gender, or party affiliation, we introduce a data-driven notion of persona grounded in collaborative filtering, which is defined as either a single individual or a cohort of individuals manifesting similar views across specific inquiries. As individuals in the same demographic group may have different personas, our data-driven persona definition allows for a more nuanced understanding of different (latent) social groups present in the population. In addition to this, we also explore an efficient method to steer LLMs toward the personas that we define. We show that our data-driven personas significantly enhance model steerability, with improvements of between $57\%-77\%$ over our best performing baselines.