Style Vectors for Steering Generative Large Language Model
Abstract
This research explores strategies for steering the output of large language models (LLMs) towards specific styles, such as sentiment, emotion, or writing style, by adding style vectors to the activations of hidden layers during text generation. We show that style vectors can be simply computed from recorded layer activations for input texts in a specific style in contrast to more complex training-based approaches. Through a series of experiments, we demonstrate the effectiveness of activation engineering using such style vectors to influence the style of generated text in a nuanced and parameterisable way, distinguishing it from prompt engineering. The presented research constitutes a significant step towards developing more adaptive and effective AI-empowered interactive systems.
Cite
@article{arxiv.2402.01618,
title = {Style Vectors for Steering Generative Large Language Model},
author = {Kai Konen and Sophie Jentzsch and Diaoulé Diallo and Peer Schütt and Oliver Bensch and Roxanne El Baff and Dominik Opitz and Tobias Hecking},
journal= {arXiv preprint arXiv:2402.01618},
year = {2024}
}
Comments
Will be published as findings paper at EACL2024 - 18th Conference of the European Chapter of the Association for Computational Linguistics