@inproceedings{lin-etal-2026-ntnu,
    title = "{NTNU}-{SMIL} at {S}em{E}val-2026 Task 3: Logistic-Loss Regression with Same-Language Transfer for Valence{--}Arousal Stance Prediction in Dimensional Stance Analysis ({D}im{S}tance)",
    author = "Lin, Siang-Ting  and
      Lo, Tien-Hong  and
      Sun, Yun-Ting  and
      Guo, Jhih-Rong  and
      Hao, Tung-Yen  and
      Tsai, Fong-Chun  and
      Chen, Berlin",
    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-1.3/",
    pages = "15--21",
    ISBN = "979-8-89176-414-9",
    abstract = "We propose NTNU-SMIL{'}s system for SemEval-2026 Task 3 Track B Subtask 1 Dimensional Stance Analysis (DimStance). Our approach models target-conditioned valence{--}arousal regression using sentence-pair encoding, dual regression heads, and a logistic-loss regression formulation. For English and Chinese, we further leverage same-language transfer from Track A and apply lightweight out-of-fold calibration with multi-seed ensembling to reduce cross-lingual scale mismatch. Post-hoc analysis shows that same-language transfer and logistic-loss regression are the main drivers of performance gains, while arousal variance collapse remains a challenge in low-resource settings such as Swahili."
}