A gentle push funziona benissimo: making instructed models in Italian via contrastive activation steering
This addresses the challenge of language adaptation for Italian tasks, offering a more efficient method than fine-tuning, though it is incremental as it builds on existing activation steering techniques.
The paper tackled the problem of adapting models to Italian, which was only partially present in pre-training data, by exploring activation steering as an alternative to fine-tuning, and found that it achieves comparable or better performance than fine-tuned models with higher quality and consistency in Italian generations.
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.