Beyond Semantics: Measuring Fine-Grained Emotion Preservation in Small Language Model-Based Machine Translation

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Wisniewski, Dawid, Czudy, Igor
Format: Preprint
Published: 2026
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866911636600651776
author Wisniewski, Dawid
Czudy, Igor
author_facet Wisniewski, Dawid
Czudy, Igor
contents Preserving affective nuance remains a challenge in Machine Translation (MT), where semantic equivalence often takes precedence over emotional fidelity. This paper evaluates the performance of three state-of-the-art Small Language Models (SLMs) -- EuroLLM, Aya Expanse, and Gemma -- in maintaining fine-grained emotions during backtranslation. Using the GoEmotions dataset, which comprises Reddit comments across 28 distinct categories, we assess emotional preservation across five European languages: German, French, Spanish, Italian, and Polish. Specifically, we investigate (i) the inherent capability of these SLMs to retain emotional sentiment, (ii) the efficacy of emotion-aware prompting in improving preservation, and (iii) the performance of ModernBERT as a contemporary alternative to BERT for emotion classification in MT evaluation.
format Preprint
id arxiv_https___arxiv_org_abs_2604_27920
institution arXiv
publishDate 2026
record_format arxiv
spellingShingle Beyond Semantics: Measuring Fine-Grained Emotion Preservation in Small Language Model-Based Machine Translation
Wisniewski, Dawid
Czudy, Igor
Computation and Language
Artificial Intelligence
Preserving affective nuance remains a challenge in Machine Translation (MT), where semantic equivalence often takes precedence over emotional fidelity. This paper evaluates the performance of three state-of-the-art Small Language Models (SLMs) -- EuroLLM, Aya Expanse, and Gemma -- in maintaining fine-grained emotions during backtranslation. Using the GoEmotions dataset, which comprises Reddit comments across 28 distinct categories, we assess emotional preservation across five European languages: German, French, Spanish, Italian, and Polish. Specifically, we investigate (i) the inherent capability of these SLMs to retain emotional sentiment, (ii) the efficacy of emotion-aware prompting in improving preservation, and (iii) the performance of ModernBERT as a contemporary alternative to BERT for emotion classification in MT evaluation.
title Beyond Semantics: Measuring Fine-Grained Emotion Preservation in Small Language Model-Based Machine Translation
topic Computation and Language
Artificial Intelligence
url https://arxiv.org/abs/2604.27920