Conspiracy theories widely spread on social media and can harm society by increasing mistrust, vaccine hesitancy, and political radicalization. However, most automated detection systems have traditionally relied on topic-specific classifiers, which often struggle to generalize across domains and provide little explanation for why a text is considered conspiratorial. To address these limitations, this paper explores various LLMs on the SemEval-2026 Task 10: psycholinguistic conspiracy marker extraction and binary conspiracy detection from Reddit submission statements. Specifically, we adopt a training-free few-shot prompting approach using different instruction-tuned large language models under a variety of few-shot settings (k in {0,1,5,10,15, 20}). Within this framework, the proposed prompting strategy incorporates psychology-informed instructions to guide the models in identifying conspiracy-related signals. As a result, the presented system achieves an F1 score of 0.21 for marker extraction and 0.81 for conspiracy detection, ranking 16th out of 30 teams in Subtask~1 and 36th out of 52 in Subtask~2 without any task-specific fine-tuning. These results suggest that psycholinguistically grounded prompting can support interpretable conspiracy analysis; however, challenges remain in identifying implicit markers.

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.

Detecting psychological defense mechanisms in supportive conversations is essential for assisting mental health practitioners. Natural language processing techniques are increasingly integral to such systems, enabling automated classification of defense levels to better understand help-seeker behavior and resistance patterns. In PsyDefDetect at BioNLP 2026, we address the task of nine-class defense level classification on the PSYDEFCONV corpus. We propose a three-stage pipeline combining LLM-based dialogue summarization, domain-specific transformer fine-tuning, and rule-based ensemble prediction. Additionally, we evaluate three mental health domain-specific transformers (Mental-BERT, Mental-RoBERTa, Mental-XLNet) alongside fine-tuned LLMs (Qwen3-4B, Qwen3-1.7B, Mistral-7B under different input conditions. Experimental results on the released test-set gold labels show that our ensemble approach achieves the best performance, reaching 34.69% macro F1 and surpassing the baseline by 4.69 percentage points. On the official PsyDefDetect Leaderboard 1 (labels 1–8), the submitted system achieved a Macro-F1 score of 23.46%, ranking 15th out of 21 teams, while on Leaderboard 2 (labels 0–8), it achieved 30.04%, securing 14th place. These findings demonstrate that domain-specific transformers substantially outperform generic LLM fine-tuning on this specialized clinical task.

Memes are now a common means of communication on social media. Their humor and short format help messages spread quickly and easily. Propagandistic memes use both words and images to influence opinions and behaviors, often appealing to emotions or ideologies. While propaganda detection has been well-studied in high-resource languages (HRLs), there has been a limited focus on low-resource languages (LRLs), such as Bengali. In this study, we introduce MemeGuard, a new dataset of 3,745 memes for detecting propaganda in Bengali. We tested more than 45 different methods, including both single and combined approaches with fusion. For text, BanglaBERT-1 achieved the best macro F1 score of 80.34%, whereas the CLIP vision transformer scored 78.94% for images. The proposed multimodal model, which combines BanglaBERT-2 and CLIP using Adaptive Modality Fusion, achieved the highest macro F1 score of 85.36%. This work establishes a strong baseline and offers valuable insights for future research in Bengali multimodal content analysis.

Large Language Models (LLMs) demonstrate exceptional proficiency in general-purpose tasks but struggle with numerical reasoning, particularly in low-resource languages like Bengali. Despite advancements, limited research has explored their numerical reasoning capabilities in these languages. To address this gap, we present BenNumEval (Bengali Numerical Evaluation), a benchmark designed to assess LLMs on numerical reasoning tasks in Bengali. It comprises six diverse tasks and a total of 3.2k samples curated from real-world problem-solving scenarios. Our extensive evaluations reveal that even with advanced prompting techniques such as Cross-Lingual Prompting (XLP) and Cross-Lingual Chain-of-Thought Prompting (XCoT), LLMs fall notably short of human-level performance, particularly when using Bengali Native Prompting (BNaP). These findings underscore the substantial gap between current LLM capabilities and human expertise in numerical reasoning, highlighting the need for more robust and linguistically inclusive AI models to advance Bengali Language Processing and equitable AI development. The source code for the system and evaluation pipeline is publicly available on GitHub.

Misogyny memes, a form of digital content, reflect societal prejudices by discriminating against women through shaming and stereotyping. In this study, we present a multimodal approach combining Indic-BERT and ViT-base-patch16-224 to address misogyny memes. We explored various Machine Learning, Deep Learning, and Transformer models for unimodal and multimodal classification using provided Tamil and Malayalam meme dataset. Our findings highlight the challenges traditional ML and DL models face in understanding the nuances of Dravidian languages, while emphasizing the importance of transformer models in capturing these complexities. Our multimodal method achieved F1-scores of 77.18% and 84.11% in Tamil and Malayalam, respectively, securing 6th place for both languages among the participants.

Fake news detection is a critical problem in today’s digital age, aiming to classify intentionally misleading or fabricated news content. In this study, we present a transformer and LLM-based ensemble method to address the challenges in fake news detection. We explored various machine learning (ML), deep learning (DL), transformer, and LLM-based approaches on a Malayalam fake news detection dataset. Our findings highlight the difficulties faced by traditional ML and DL methods in accurately detecting fake news, while transformer- and LLM-based ensemble methods demonstrate significant improvements in performance. The ensemble method combining Sarvam-1, Malayalam-BERT, and XLM-R outperformed all other approaches, achieving an F1-score of 89.30% on the given dataset. This accomplishment, which contributed to securing 2nd place in the shared task at DravidianLangTech 2025, underscores the importance of developing effective methods for detecting fake news in Dravidian languages.

Detecting hate speech in Bengali social media content presents considerable challenges, primarily due to the prevalence of informal language and the limited availability of annotated datasets. This study investigates the identification of hate speech in Bengali YouTube comments, focusing on classifying the type, severity, and target group. Multiple machine learning baselines and voting ensemble techniques are evaluated to address these tasks. The methodology involves text preprocessing, feature extraction using TF-IDF and Count vectors, and aggregating predictions from several models. Hierarchical classification with TF-IDF features and majority voting improves the detection of less frequent hate speech categories while maintaining robust overall performance, resulting in an 18^th place ranking and a micro F1 score of 68.42%. Furthermore, ablation studies assess the impact of preprocessing steps and n-gram selection, providing reproducible baselines for Bengali hate speech detection. All codes and resources are publicly available at https://github.com/Hasan-Mesbaul-Ali-Taher/BLP_25_Task_1

Hate speech on social media platforms, particularly in low-resource languages like Bengali, poses a significant challenge due to its nuanced nature and the need to understand its type, severity, and targeted group. To address this, the Bangla Multi-task Hate Speech Identification Shared Task at BLP 2025 adopts a multi-task learning framework that requires systems to classify Bangla YouTube comments across three subtasks simultaneously: type of hate, severity, and targeted group. To tackle these challenges, this work presents BanTriX, a transformer ensemble method that leverages BanglaBERT-I, XLM-R, and BanglaBERT-II. Evaluation results show that the BanTriX, optimized with cross-entropy loss, achieves the highest weighted micro F1-score of 73.78% in Subtask 1C, securing our team 2nd place in the shared task.

Distinguishing fact from opinion in text is a nuanced but essential task, particularly in news articles where subjectivity can influence interpretation and reception. Identifying whether content is subjective or objective is critical for sentiment analysis, media bias detection, and content moderation. However, progress in this area has been limited for low-resource languages such as Bengali due to a lack of benchmark datasets and tools. To address these constraints, this work presents BeNSD (Bengali News Subjectivity Detection), a novel dataset of 8,655 Bengali news article texts, along with an enhanced transformer-based architecture (POS-Aware-MuRIL) that integrates parts-of-speech (POS) features with MuRIL embeddings at the input level to provide richer contextual representation for subjectivity detection. A range of baseline models is evaluated, and the proposed architecture achieves a macro F1-score of 93.35% in subjectivity detection for the Bengali language.

In this modern era, many people have been using Facebook and Twitter, leading to increased information sharing and communication. However, a considerable amount of information on these platforms is misleading or intentionally crafted to deceive users, which is often termed as fake news. A shared task on fake news detection in Malayalam organized by DravidianLangTech@EACL 2024 allowed us for addressing the challenge of distinguishing between original and fake news content in the Malayalam language. Our approach involves creating an intelligent framework to categorize text as either fake or original. We experimented with various machine learning models, including Logistic Regression, Decision Tree, Random Forest, Multinomial Naive Bayes, SVM, and SGD, and various deep learning models, including CNN, BiLSTM, and BiLSTM + Attention. We also explored Indic-BERT, MuRIL, XLM-R, and m-BERT for transformer-based approaches. Notably, our most successful model, m-BERT, achieved a macro F1 score of 0.85 and ranked 4th in the shared task. This research contributes to combating misinformation on social media news, offering an effective solution to classify content accurately.

Violence-inciting text detection has become critical due to its significance in social media monitoring, online security, and the prevention of violent content. Developing an automatic text classification model for identifying violence in languages with limited resources, like Bangla, poses significant challenges due to the scarcity of resources and complex morphological structures. This work presents a transformer-based method that can classify Bangla texts into three violence classes: direct, passive, and non-violence. We leveraged transformer models, including BanglaBERT, XLM-R, and m-BERT, to develop a hierarchical classification model for the downstream task. In the first step, the BanglaBERT is employed to identify the presence of violence in the text. In the next step, the model classifies stem texts that incite violence as either direct or passive. The developed system scored 72.37 and ranked 14th among the participants.