@inproceedings{aslam-etal-2026-team,
    title = "Team Habib Disambiguators at {S}em{E}val-2026 Task 5: Assessing Semantic Plausibility using Regularized Transformer Fine-Tuning",
    author = "Aslam, Zohaib  and
      Siddiqui, Ahsan  and
      Enayet, Ayesha",
    editor = "Kochmar, Ekaterina  and
      Ghosh, Debanjan  and
      North, Kai  and
      Komachi, Mamoru",
    booktitle = "Proceedings of the 20th {I}nternational {W}orkshop on {S}emantic {E}valuation (2026)",
    month = jul,
    year = "2026",
    address = "San Diego, California, USA",
    publisher = "Association for Computational Linguistics",
    url = "https://preview.aclanthology.org/ingest-acl-workshops/2026.semeval-1.269/",
    pages = "2126--2129",
    ISBN = "979-8-89176-414-9",
    abstract = "This paper presents a system for SemEval-2026 Task 5: Rating Plausibility of Word Senses in Ambiguous Sentences through Narrative Understanding. The task involves predicting the plausibility of a specific word sense within a short story where context provided by the ending resolves a deliberate ambiguity. We model this as a regression problem, fine-tuning a DeBERTa-v3 transformer to predict the distribution of human judgments rather than a single hard label. To address the challenge of limited training data and potential overfitting, we employ R-Drop (Consistency Regularization) to enforce prediction stability across dropout masks and Layer-wise Learning Rate Decay (LLRD) to preserve the model{'}s pre-trained linguistic knowledge. Our experiments demonstrate that treating plausibility as a soft-label distribution, combined with aggressive regularization, improves generalization on ambiguous samples. The submitted system achieves a Spearman correlation of 0.56 and an Accuracy (within SD) of 0.74 on the official test set."
}