Work

Lost in Activations: A Neuron-level Analysis of Encoders for Cross-Lingual Emotion Detection

Year: 2026

Type: conference-paper

Abstract: The rapid advancement of multilingual pre-trained transformers has fueled significant progress in natural language understanding across diverse languages.Yet, their inner workings remain opaque, especially with regard to how individual neurons encode and generalize semantic and affective features across languages.This paper presents an interpretability study of a fine-tuned XLM-R model for multilingual emotion classification.Using neuron-level activation analysis, we investigate the variance of neurons across labels, cross-lingual alignment of activations, and the existence of "polyglot" versus language-specific neurons.Our results reveal that while certain neurons consistently encode emotion-related concepts across languages, others show strong monolingual specialization.The code for the analysis is publicly available at https://github.com/pranaydeeps/emotion_interp.

Source: Conference of the European Chapter of the Association for Computational Linguistics

DOI: 10.18653/v1/2026.eacl-short.9

Authors: Pranaydeep Singh, Orphee De Clercq, Els Lefever

Institution: Ghent University

Language: English

Topic: Explainable Artificial Intelligence (XAI)

Subfield: Artificial Intelligence

Field: Computer Science

Domain: Physical Sciences

Open Access status: hybrid