SubmissionNumber#=%=#91
FinalPaperTitle#=%=#UniBuc at SemEval-2024 Task 2: Tailored Prompting with Solar for Clinical NLI
ShortPaperTitle#=%=#
NumberOfPages#=%=#10
CopyrightSigned#=%=#Claudiu Creanga
JobTitle#==#
Organization#==#
Abstract#==#This paper describes the approach of the UniBuc team in tackling the SemEval 2024 Task 2: Safe Biomedical Natural Language Inference for Clinical Trials. We used SOLAR Instruct, without any fine-tuning, while focusing on input manipulation and tailored prompting. By customizing prompts for individual CTR sections, in both zero-shot and few-shots settings, we managed to achieve a consistency score of 0.72, ranking 14th in the leaderboard. Our thorough error analysis revealed that our model has a tendency to take shortcuts and rely on simple heuristics, especially when dealing with semantic-preserving changes.
Author{1}{Firstname}#=%=#Marius
Author{1}{Lastname}#=%=#Micluta-Campeanu
Author{1}{Username}#=%=#marius.micluta-campeanu
Author{1}{Email}#=%=#marius.micluta-campeanu@my.fmi.unibuc.ro
Author{1}{Affiliation}#=%=#University of Bucharest
Author{2}{Firstname}#=%=#Claudiu
Author{2}{Lastname}#=%=#Creanga
Author{2}{Username}#=%=#claudiucreanga
Author{2}{Email}#=%=#claudiu.creanga.backup@gmail.com
Author{2}{Affiliation}#=%=#University of Bucharest
Author{3}{Firstname}#=%=#Ana-Maria
Author{3}{Lastname}#=%=#Bucur
Author{3}{Username}#=%=#bucuram
Author{3}{Email}#=%=#ana-maria.bucur@drd.unibuc.ro
Author{3}{Affiliation}#=%=#Interdisciplinary School of Doctoral Studies
Author{4}{Firstname}#=%=#Ana Sabina
Author{4}{Lastname}#=%=#Uban
Author{4}{Username}#=%=#anauban
Author{4}{Email}#=%=#ana.uban@gmail.com
Author{4}{Affiliation}#=%=#University of Bucharest
Author{5}{Firstname}#=%=#Liviu P.
Author{5}{Lastname}#=%=#Dinu
Author{5}{Username}#=%=#liviu.p.dinu
Author{5}{Email}#=%=#liviu.p.dinu@gmail.com
Author{5}{Affiliation}#=%=#University of Bucharest

==========
èéáğö