Sharing Matters: Analysing Neurons Across Languages and Tasks in LLMs
Abstract: Multilingual large language models (LLMs) have greatly increased the ceiling of performance on non-English tasks. However, the mechanisms behind multilingualism in these LLMs are poorly understood. Of particular interest is the degree to which internal representations are shared between languages. Recent work on neuron analysis of LLMs has focused on the monolingual case, and the limited work on the multilingual case has not considered the interaction between tasks and linguistic representations. In our work, we investigate how neuron activation is shared across languages by categorizing neurons into four distinct groups according to their responses across different languages for a particular input: all-shared, partial-shared, specific, and non-activated. This categorization is combined with a study of neuron attribution, i.e. the importance of a neuron w.r.t an output. Our analysis reveals the following insights: (i) the linguistic sharing patterns are strongly affected by the type of task, but neuron behavior changes across different inputs even for the same task; (ii) all-shared neurons play a key role in generating correct responses; (iii) boosting multilingual alignment by increasing all-shared neurons can enhance accuracy on multilingual tasks. We will release the code to foster research in this area.
Paper Type: Long
Research Area: Multilingualism and Cross-Lingual NLP
Research Area Keywords: multilingual interpretability, multilingual representations, neuron behavior, feature attribution
Contribution Types: Model analysis & interpretability
Languages Studied: English, German, Spanish, French, Portuguese, Russian, Thai, Turkish, Vietnamese, Chinese
Submission Number: 2504
Loading