We describe a unified system for SemEval-2026 Task 9 on multilingual polarization detection. The task requires binary polarization detection, multi-label target type classification, and multi-label manifestation identification across languages and events with severe class imbalance. Our approach combines (i) targeted data augmentation for low-frequency labels, (ii) merged multitask fine-tuning of Subtask 2 and Subtask 3, and (iii) model fusion to improve cross-lingual stability. Subtask 1 predictions are derived via calibrated inference from the multi-label head. On the development set, multitask training consistently out-performs single-task variants, and fusion yields additional gains, especially for rare labels. We also report ablations and error analyses, highlighting remaining challenges such as implicit polarization and partial-label uncertainty.

Large language models (LLMs) often exhibit significant cultural representation biases in multilingual everyday knowledge understanding, struggling to accurately capture region-specific customs and values. This paper presents our system submission for SemEval 2026 Task 7: BLEnD Challenge Track 2 (MCQ) (SemEval-2026 Task 7 Organizers, 2026). To address these challenges, we propose a training-free retrieval-augmented generation (RAG) framework. Without introducing any external data, we manuallyconstructed a localized multicultural knowledge base for each language-region and used text-embedding-v4 for region-specific cultural background retrieval. In the generation stage, we adopted a strict zero-shot setting: prompts contain no task instance question-answer examples, only injecting locale-relevant background cultural descriptions via RAG to compensate for contextual information absence, combined with a dual-model ensemble strategy using Gemini 3 Flash (preview) (Google DeepMind, 2025) and GPT-5.2 Chat (OpenAI, 2025). Our system achieved an overall score of 96.35 on the final Evaluation dataset.Additionally, we conducted in-depth analysis of model performance on specific languages, particularly highlighting severe cultural alignment challenges faced by large models in dialectal variants like Moroccan Arabic (ar-MA) and highly localized subjective Japanese (jaJP) everyday scenarios

This paper describes our system used in the SemEval-2026 Task 7: Cross-Language Cultural Everyday Knowledge QA (track 1). Cultural knowledge typically exhibits significant regional specificity and is deeply rooted in particular linguistic conventions, posing severe challenges to general-purpose large language models (LLMs). We propose a retrieval-augmented generation (RAG) framework: this framework utilizes text-embedding-v4 as the retrieval core to precisely extract social knowledge and expression patterns from region-specific large-scale multilingual cultural knowledge bases, and drives the gpt-5.2-chat model to generate concise answers that are both logically factual and highly aligned with the target region’s cultural context. In the official evaluation, our system ranked first among all participating teams with a total score of 78.7672, fully demonstrating the method’s outstanding performance in cross-cultural accuracy and linguistic authenticity.