Online polarization, defined as the pronounced division of public opinion into antagonistic groups, poses a significant threat to social cohesion. Automatic detection of polarization across diverse languages and cultures is essential for effective monitoring of online discourse. The challenge extends beyond identifying hate speech to recognizing more nuanced forms, including negative stereotypes, attribution of blame, and dehumanization. This work addresses SemEval-2026 Task 9, which focuses on detecting polarization in multiple languages. Specifically, Subtask 1 involves binary classification of message polarization, while Subtask 2 requires assigning multiple polarization labels in English and Bengali. For Subtask 1, Qwen3-14B is employed with structured few-shot prompting in 4-bit mode, yielding test macro-F1 scores of 0.847 for Bengali (4th place) and 0.808 for English (9th place). For Subtask 2, XLM-RoBERTa-large and RoBERTa-base are fine-tuned using an uneven loss (γ+ = 1, γ− =4) and label-specific thresholds, which increase development macro F1 by up to 24.6 points. The final test macro F1 for English is 0.454 (21st place). Analysis indicates that large language model prompting enhances binary polarization detection, while threshold adjustment is critical for addressing class imbalance in multi-label tasks.

Detecting equivocation is essential, as indirect or evasive responses can shape public perception, influence political narratives, and undermine transparency in democratic discourse. To address the challenge of detecting evasive political responses on digital platforms, participation in the CLARITY SemEval-2026 Task was undertaken, which focuses on (i) clarity-level classification and (ii) fine-grained evasion-type classification in political question-answer contexts. This study introduces a data-centric framework that systematically examines the effects of class distribution and refinement strategies on the performance of Large Language Models (LLMs). A distribution-aware, LLM-augmented dataset was constructed by selectively paraphrasing minority-class instances to enhance class balance, and its performance was benchmarked against full, rebalanced, and undersampled training configurations. To comprehensively assess the proposed method, Qwen3-14B, Phi-4, Gemma-2 9B, and Mistral 7B were evaluated in in-context learning (ICL) settings (zero-shot and few-shot) and with LoRA fine-tuning. Experimental results indicate that fine-tuning Phi-4 with class rebalancing yields strong performance, achieving 74.77% on Subtask-1 and 51.55% on Subtask-2. Consequently, the system ranked 21st in Subtask-1 and 22nd in Subtask-2 on the official evaluation leaderboard.

Ensuring a safe and inclusive online environment requires effective hate speech detection on social media. While detection systems have significantly advanced for English, many regional languages, including Malayalam, Tamil and Telugu, remain underrepresented, creating challenges in identifying harmful content accurately. These languages present unique challenges due to their complex grammar, diverse dialects, and frequent code-mixing with English. The rise of multimodal content, including text and audio, adds further complexity to detection tasks. The shared task “Multimodal Hate Speech Detection in Dravidian Languages: DravidianLangTech@NAACL 2025” has aimed to address these challenges. A Youtube-sourced dataset has been provided, labeled into five categories: Gender (G), Political (P), Religious (R), Personal Defamation (C) and Non-Hate (NH). In our approach, we have used mBERT, T5 for text and Wav2Vec2 and Whisper for audio. T5 has performed poorly compared to mBERT, which has achieved the highest F1 scores on the test dataset. For audio, Wav2Vec2 has been chosen over Whisper because it processes raw audio effectively using self-supervised learning. In the hate speech detection task, we have achieved a macro F1 score of 0.2005 for Malayalam, ranking 15th in this task, 0.1356 for Tamil and 0.1465 for Telugu, with both ranking 16th in this task.

Critical Question Generation (CQs-Gen) improves reasoning and critical thinking skills through Critical Questions (CQs), which identify reasoning gaps and address misinformation in NLP, especially as LLM-based chat systems are widely used for learning and may encourage superficial learning habits. The Shared Task on Critical Question Generation, hosted at the 12th Workshop on Argument Mining and co-located in ACL 2025, has aimed to address these challenges. This study proposes a CQs-Gen pipeline using Llama-3-8B-Instruct-GGUF-Q8_0 with few-shot learning, integrating text simplification, NER, and argument schemes to enhance question quality. Through an extensive experiment testing without training, fine-tuning with PEFT using LoRA on 10% of the dataset, and few-shot fine-tuning (using five examples) with an 8-bit quantized model, we demonstrate that the few-shot approach outperforms others. On the validation set, 397 out of 558 generated CQs were classified as Useful, representing 71.1% of the total. In contrast, on the test set, 49 out of 102 generated CQs, accounting for 48% of the total, were classified as Useful following evaluation through semantic similarity and manual assessments.