Paper Detail
Srikar Kashyap Pulipaka
We present our system for SemEval-2026 Task 9: Multilingual Polarization Detection, a binary classification task spanning 22 languages. Our approach fine-tunes separate Gemma~3 models (12B and 27B parameters) per language using Low-Rank Adaptation (LoRA), augmented with synthetic data generated by a large language model (LLM). We employ three synthetic data strategies (direct generation, paraphrasing, and contrastive pair creation) using GPT-4o-mini, with a multi-stage quality filtering pipeline including embedding-based deduplication. We find that per-language threshold tuning on the development set yields 2 to 4\% F1 improvements without retraining. We also use weighted ensembles of 12B and 27B model predictions with per-language strategy selection. Our final system achieves a mean macro-F1 of 0.811 across all 22 languages, ranking 2nd overall of the participating teams, with 1st place finishes in 3 languages and top-3 in 8 languages. We also find that alternative architectures (XLM-RoBERTa, Qwen3) that showed strong development set performance suffered 30 to 50\% F1 drops on the test set, highlighting the importance of generalization.
No structured notes yet. Add `summary_sections`, `why_relevant`, `claim_impact`, or `next_action` in `papers.jsonl` to enrich this view.
No ranking explanation is available yet.
No tags.
@article{pulipaka2026psk,
title = {PSK at SemEval-2026 Task 9: Multilingual Polarization Detection Using Ensemble Gemma Models with Synthetic Data Augmentation},
author = {Srikar Kashyap Pulipaka},
year = {2026},
abstract = {We present our system for SemEval-2026 Task 9: Multilingual Polarization Detection, a binary classification task spanning 22 languages. Our approach fine-tunes separate Gemma\textasciitilde{}3 models (12B and 27B parameters) per language using Low-Rank Adaptation (LoRA), augmented with synthetic data generated by a large language model (LLM). We employ three synthetic data strategies (direct generation, paraphrasing, and contrastive pair creation) using GPT-4o-mini, with a multi-stage quality filtering pipeline},
url = {https://arxiv.org/abs/2605.05159},
keywords = {cs.CL, cs.AI, cs.LG},
eprint = {2605.05159},
archiveprefix = {arXiv},
}
{}