SubmissionNumber#=%=#8
FinalPaperTitle#=%=#Otter at MedExAct2026: Diverse Encoder Ensemble for Medical Decision Span Detection
ShortPaperTitle#=%=#
NumberOfPages#=%=#
CopyrightSigned#=%=#
JobTitle#==#
Organization#==#
Abstract#==#We build an ensemble of 10 transformer encoders for the MedExACT 2026 shared task on medical decision span detection. The ensemble is diversified along three training directions: encoder initialization (including domain-adaptive pre-training on clinical text), loss function, and data augmentation with LLM-generated synthetic notes and silver-labeled clinical documents. Greedy forward search selects the combination with the highest validation final score. A BERT-based boundary refiner is applied to the ensemble's predicted spans to correct offset errors before submission.
Author{1}{Firstname}#=%=#Lalita
Author{1}{Lastname}#=%=#Lowphansirikul
Author{1}{Username}#=%=#lalital
Author{1}{Orcid}#=%=#
Author{1}{Email}#=%=#lalita.lowphansirikul@gmail.com
Author{1}{Affiliation}#=%=#Vidyasirimedhi Institute of Science and Technology
Author{2}{Firstname}#=%=#Piyalitt
Author{2}{Lastname}#=%=#Ittichaiwong
Author{2}{Username}#=%=#piyalitt.itt
Author{2}{Orcid}#=%=#
Author{2}{Email}#=%=#piyalitt.itt@mahidol.ac.th
Author{2}{Affiliation}#=%=#Faculty of Medicine Siriraj Hospital, Mahidol University

==========
èéáğö