ReCoVeR the Target Language: Language Steering without Sacrificing Task Performance
Published in Findings of the Association for Computational Linguistics: EMNLP 2025, 2025
As they become increasingly multilingual, Large Language Models exhibit more language confusion, i.e., they tend to generate answers in a language different from the language of the prompt or the answer language explicitly requested by the user. In this work, we propose ReCoVeR (REducing language COnfusion in VEctor Representations), a novel lightweight approach for reducing language confusion based on language-specific steering vectors.