A gentle push funziona benissimo: making instructed models in Italian via contrastive activation steering

TL;DR

This paper investigates activation steering as a cost-effective alternative to fine-tuning for improving Italian language model performance, demonstrating comparable or superior results and higher generation quality.

Contribution

It introduces activation steering techniques tailored for Italian, showing their effectiveness across different models without extensive fine-tuning.

Findings

Activation steering improves Italian task performance.

Steering achieves comparable or better results than fine-tuning.

Higher quality and consistency in Italian generations.

Abstract

Adapting models to a language that was only partially present in the pre-training data requires fine-tuning, which is expensive in terms of both data and computational resources. As an alternative to fine-tuning, we explore the potential of activation steering-based techniques to enhance model performance on Italian tasks. Through our experiments we show that Italian steering (i) can be successfully applied to different models, (ii) achieves performances comparable to, or even better than, fine-tuned models for Italian, and (iii) yields higher quality and consistency in Italian generations. We also discuss the utility of steering and fine-tuning in the contemporary LLM landscape where models are anyway getting high Italian performances even if not explicitly trained in this language.