In-context Learning vs. Instruction Tuning: The Case of Small and Multilingual Language Models
This work addresses the problem of extending instruction following to diverse languages and model sizes for researchers and practitioners, but it is incremental as it compares existing methods without introducing new ones.
The study compared in-context learning and instruction tuning for instruction following in small and multilingual language models across English, French, and Spanish, finding that multilingual and smaller models led to reduced in-context learning performance, with only partial improvement from Direct Preference Optimisation alignment.
Instruction following is a critical ability for Large Language Models to perform downstream tasks. The standard approach to instruction alignment has relied on a specific phase of model tuning over curated instruction datasets, optionally complemented with an alignment step over human preferences. Recent work has shown the potential of in-context learning (ICL) alternatives to guide base models towards instruction following. This type of approach is particularly relevant to extend instruction following across languages and models of varying sizes adapted to different types of usage. In this work we compare ICL and instruction fine-tuning in English, French and Spanish, on Small Language Models, and provide experimental results on applying Direct Preference Optimisation (DPO) over base models. Our results show that scenarios involving multilingual and smaller models result in downgraded ICL instruction following performance, only partially mitigated by DPO alignment. This study aims to further our understanding of current strengths and limitations of alternative methods for instruction following.