Fine-tuning XLM-RoBERTa-base with separate models per language-domain pair outperforms few-shot LLMs for multilingual dimensional aspect sentiment regression.
but the staff was so horrible to us
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.CL 1years
2026 1verdicts
UNVERDICTED 1representative citing papers
citing papers explorer
-
NCL-BU at SemEval-2026 Task 3: Fine-tuning XLM-RoBERTa for Multilingual Dimensional Sentiment Regression
Fine-tuning XLM-RoBERTa-base with separate models per language-domain pair outperforms few-shot LLMs for multilingual dimensional aspect sentiment regression.