NCL-UoR at SemEval-2026 Task 5: Embedding-Based Methods, Fine-Tuning, and LLMs for Word Sense Plausibility Rating

NCL-UoR at SemEval-2026 Task 5: Embedding-Based Methods, Fine-Tuning, and LLMs for Word Sense Plausibility Rating explores A structured prompting strategy for word sense plausibility rating that outperforms fine-tuned models and embedding-based approaches, offering a calibrated rating system for ambiguous homonyms in narrative contexts.. Commercial viability score: 7/10 in NLP.