Large language models are increasingly used in applications where alignment with human values is critical. While model fine-tuning is often employed to ensure safe responses, this technique is static and does not lend itself to everyday situations involving dynamic values and preferences. In this paper, we present a practical, reproducible, and model-agnostic procedure to evaluate whether a prompt candidate can effectively steer generated text toward specific human values, formalising a scoring method to quantify the presence and gain of target values in generated responses. We apply our method to a variant of the Wizard-Vicuna language model, using Schwartz's theory of basic human values and a structured evaluation through a dialogue dataset. With this setup, we compare a baseline prompt to one explicitly conditioned on values, and show that value steering is possible even without altering the model or dynamically optimising prompts.
@article{arxiv.2511.16688,
title = {Prompt-Based Value Steering of Large Language Models},
author = {Giulio Antonio Abbo and Tony Belpaeme},
journal= {arXiv preprint arXiv:2511.16688},
year = {2025}
}
Comments
9 pages, 1 figure, 4 tables. Presented at the 3rd International Workshop on Value Engineering in AI (VALE 2025), 28th European Conference on AI. To appear in Springer LNCS