Seals-NLP at SemEval-2026 Task 9: A Comparative Study of Transformer Architectures for Polarization Detection
Year: 2026
Type: conference-paper
Abstract: We describe the Seals-NLP system for SemEval-2026 Task 9 (POLAR) Subtask 1, binary polarization detection.Our study compares (i) fully fine-tuned encoder-only transformers, (ii) QLoRA-based fine-tuned openweight LLMs, and (iii) zero-shot prompted LLMs.ModernBERT-large emerges as the most cost-effective option, matching or surpassing larger fine-tuned and zero-shot LLMs in macro-F 1 while requiring substantially less memory and lower latency.An error analysis by failure mode and polarization subtype reveals systematic over-triggering on political cue words and under-detection of sarcastic vilification and multifaceted attacks in the POLAR dataset across all models.
Authors: Minh Smith, Cheryl Seals
Institution: Auburn University
Language: English
Subfield: Artificial Intelligence
Field: Computer Science
Domain: Physical Sciences
Open Access status: hybrid
OpenAlex ID: W7166877552