---
title: "CYUT at SemEval-2026 Task 9 — paper by Wu et al. (2026)"
description: "The CYUT team tackles SemEval-2026 Task 9, a benchmark of online polarization across 22 languages… 12 references, cited by 1 paper here."
authors: ["Shih-Hung Wu", "Yun-Kuang Liao", "Shih-Siang Su", "Yi-Min Jian"]
year: 2026
venue: "Proceedings of the 20th International Workshop on Semantic Evaluation (2026)"
source: "https://aclanthology.org/2026.semeval-1.209/"
doi: "10.18653/v1/2026.semeval-1.209"
licence: "CC BY 4.0 (ACL Anthology)"
licence_url: "https://creativecommons.org/licenses/by/4.0/"
summary_licence: "Elamur Cosmic License"
canonical: "https://research.elamur.ai/free/en/papers/wu-2026-cyut-semeval-2026-task-9/"
---
[Elamur Research](https://md.research.elamur.ai/free/en/) · [this page in HTML](https://research.elamur.ai/free/en/papers/wu-2026-cyut-semeval-2026-task-9/)

[home](https://md.research.elamur.ai/free/en/) / [papers](https://md.research.elamur.ai/free/en/papers/) / [2026](https://md.research.elamur.ai/free/en/papers/2026/page/11/) / CYUT at SemEval-2026 Task 9: Monolingual vs. Multilingual LoRA Tuning for Multicultural and Multievent Polarization Detection

# CYUT at SemEval-2026 Task 9: Monolingual vs. Multilingual LoRA Tuning for Multicultural and Multievent Polarization Detection

**Authors:** [Shih-Hung Wu](https://md.research.elamur.ai/free/en/authors/shih-hung-wu/), Yun-Kuang Liao, Shih-Siang Su and [Yi-Min Jian](https://md.research.elamur.ai/free/en/authors/yi-min-jian/) · **Year:** 2026 · **Venue:** [Proceedings of the 20th International Workshop on Semantic Evaluation (2026)](https://md.research.elamur.ai/free/en/venues/semeval/) ([SemEval 2026](https://md.research.elamur.ai/free/en/venues/semeval/2026/)) · **Publisher:** [Association for Computational Linguistics](https://md.research.elamur.ai/free/en/publishers/association-for-computational-linguistics/) · **Licence:** [CC BY 4.0 (ACL Anthology)](https://creativecommons.org/licenses/by/4.0/)

**Links:** [Publisher's page](https://aclanthology.org/2026.semeval-1.209/) · [PDF](https://aclanthology.org/2026.semeval-1.209.pdf) · [DOI 10.18653/v1/2026.semeval-1.209](https://doi.org/10.18653/v1/2026.semeval-1.209) · [OpenAlex](https://openalex.org/W7166822746)

**Citations:** 0 citations (OpenAlex, 2026-10-14) · [1](https://md.research.elamur.ai/free/en/papers/wu-2026-cyut-semeval-2026-task-9/cited-by/) in our library, of the 149,439 papers whose bibliographies we read

## Summary
The CYUT team tackles SemEval-2026 Task 9, a benchmark of online polarization across 22 languages with three subtasks: binary polarization detection, classification of polarization type (such as political or religious targets), and identification of rhetorical manifestations (such as stereotyping or dehumanization). They LoRA-fine-tune two open instruction-tuned LLMs, Llama-3.1-8B-Instruct and Qwen2-7B-Instruct, using two data setups:

- **Multilingual, task-separated:** one model per subtask, trained on all languages pooled together.
- **Chinese-only, multi-task:** a single model trained on Chinese data with the labels of all three subtasks combined.

They then merge three such systems in an ensemble weighted by validation macro-F1. The three systems were chosen to differ deliberately for diversity, not as controlled comparisons. The ensemble was their official submission and ranked first for Chinese in Subtask 1. It beat the official baseline in several languages but not consistently in all. The Chinese-only model is strong on Chinese but much weaker on other languages. The work shows that combining monolingual and multilingual adapters can help in a high-resource target language.

## Key points
- **Single systems:**
  - M1 = Llama-3.1-8B-Instruct, multilingual and task-specific.
  - M2 = Qwen2-7B-Instruct with the same setup as M1.
  - M3 = Qwen2-7B-Instruct, Chinese-only with joint tasks.
- **Ensemble (M4):** combines M1–M3 with normalized weights. Subtask 1 uses per-language macro-F1 as weights. Subtasks 2 and 3 use per-label macro-F1 weights.
- **Chinese (zho) results, Subtask 1:** M4 reached macro-F1 0.9315, compared with 0.8691 for the baseline and 0.9248 for M3.
- **Chinese (zho) results, Subtasks 2 and 3:**
  - Subtask 2: M3 scored 0.8262, M4 scored 0.8245, and the baseline scored 0.6697.
  - Subtask 3: M4 scored 0.7004.
- **Results across languages:**
  - M1 had the best overall results on Subtasks 1 and 3; M2 was best on Subtask 2.
  - M3 often collapsed outside Chinese, falling to about 0.11 on one language in Subtask 1.
  - The baseline remained stronger in many Subtask 3 languages.

*Our summary of the paper, in our own words — not its text. The work belongs to its authors; read it at the links above.*

## Institutions

- [Chaoyang University of Technology](https://md.research.elamur.ai/free/en/universities/chaoyang-university-of-technology/) — University, Taichung, Taiwan · 2 works · 2 citations

📚 Shelves (4): [🖥️ Computer Science](https://md.research.elamur.ai/free/en/shelves/computer-science/) · 16,667 works · [🤖 Artificial Intelligence](https://md.research.elamur.ai/free/en/shelves/computer-science/artificial-intelligence/) · 16,194 works · [🎭 Arts and Humanities](https://md.research.elamur.ai/free/en/shelves/arts-and-humanities/) · 16,039 works · [🗣️ Language and Linguistics](https://md.research.elamur.ai/free/en/shelves/arts-and-humanities/language-and-linguistics/) · 15,891 works

🧭 Topics (1): [🌍 Multilingual & cross-lingual NLP](https://md.research.elamur.ai/free/en/topics/multilingual-nlp/) · 828 works

## Cite

Cite as: Shih-Hung Wu, Yun-Kuang Liao, Shih-Siang Su and Yi-Min Jian. 2026. CYUT at SemEval-2026 Task 9: Monolingual vs. Multilingual LoRA Tuning for Multicultural and Multievent Polarization Detection. Proceedings of the 20th International Workshop on Semantic Evaluation (2026).

The paper itself, not this page. BibTeX from the library's record of the work.

```bibtex
@inproceedings{wu2026cyut,
  title = {{CYUT at SemEval-2026 Task 9: Monolingual vs. Multilingual LoRA Tuning for Multicultural and Multievent Polarization Detection}},
  author = {Shih-Hung Wu, Yun-Kuang Liao, Shih-Siang Su and Yi-Min Jian},
  booktitle = {Proceedings of the 20th International Workshop on Semantic Evaluation (2026)},
  year = {2026},
  doi = {10.18653/v1/2026.semeval-1.209},
  url = {https://doi.org/10.18653/v1/2026.semeval-1.209}
}
```

- [← previous CYUT at SemEval-2026 Task 3: Multi-Task Dimensional Aspect Sentiment Regression with Polar Multi-Zone Labeling in VA Space](https://md.research.elamur.ai/free/en/papers/wu-2026-cyut-semeval-2026-task-3/)
- [next → d-itlab at SemEval-2026 Task 12: Per-Option Surprisal and Multi-Stage Gating for Precision-Oriented Causal Reasoning](https://md.research.elamur.ai/free/en/papers/terao-2026-d-itlab-semeval-2026-task/)
- [↑ back to papers of 2026](https://md.research.elamur.ai/free/en/papers/2026/page/11/)

## References

12 references, 6 of them with a card of ours. ↗ leads to the work, «find» searches for one we have no id for yet.

1. Pinzhen Chen, Shaoxiong Ji, Nikolay Bogoychev, Andrey Kutuzov, Barry Haddow, and Kenneth Heafield. 2024. Monolingual or multilingual instruction tuning: Which makes a better alpaca. In Findings of the Association for Computational Linguistics: EACL 2024, pages 1347–1356. [find](https://scholar.google.com/scholar?q=Pinzhen+Chen%2C+Shaoxiong+Ji%2C+Nikolay+Bogoychev%2C+Andrey+Kutuzov%2C+Barry+Haddow%2C+and+Kenneth+Heafield%2E+2024%2E+Monolingual+or+multilingual+instruction+tuning%3A+Which+makes+a+better+alpaca%2E+In+Findings+of+the)
2. Nick Haslam. 2006. Dehumanization: An integrative review. Personality and social psychology review, 10(3):252–264. [find](https://scholar.google.com/scholar?q=Nick+Haslam%2E+2006%2E+Dehumanization%3A+An+integrative+review%2E+Personality+and+social+psychology+review%2C+10%283%29%3A252%E2%80%93264%2E)
3. Edward J Hu, Yelong Shen, Phillip Wallis, Zeyuan Allen-Zhu, Yuanzhi Li, Shean Wang, Liang Wang, Weizhu Chen, and 1 others. 2022. [Lora: Low-rank adaptation of large language models](https://md.research.elamur.ai/free/en/papers/hu-2021-lora-low-rank-adaptation-large/). Iclr, 1(2):3.
4. Pratik Joshi, Sebastin Santy, Amar Budhiraja, Kalika Bali, and Monojit Choudhury. 2020. [The state and fate of linguistic diversity and inclusion in the nlp world](https://md.research.elamur.ai/free/en/papers/joshi-2020-state-fate-linguistic-diversity-inclusion/). In Proceedings of the 58th annual meeting of the association for computational linguistics, pages 6282–6293.
5. Usman Naseem, Robert Geislinger, Juan Ren, Sarah Kohail, Rudy Garrido Veliz, P Sam Sahil, Yiran Zhang, Marco Antonio Stranisci, Idris Abdulmumin, "Ozge Alacam, Cengiz Acar"urk, Aisha Jabr, Saba Anwar, Abinew Ali Ayele, Elena Tutubalina, Aung Kyaw Htet, Xintong Wang, Surendrabikram Thapa, Tanmoy Chak… — [SemEval-2026 Task 9: Detecting Multilingual, Multicultural and Multievent Online Polarization](https://md.research.elamur.ai/free/en/papers/naseem-2026-semeval-2026-task-9-detecting/)
6. Usman Naseem, Robert Geislinger, Juan Ren, Sarah Kohail, Rudy Garrido Veliz, P Sam Sahil, Yiran Zhang, Marco Antonio Stranisci, Idris Abdulmumin, Özge Alacam, Cengiz Acartürk, Aisha Jabr, Saba Anwar, Abinew Ali Ayele, Simona Frenda, Alessandra Teresa Cignarella, Elena Tutubalina, Oleg Rogov, Aung Ky… — [POLAR: A Benchmark for Multilingual, Multicultural, and Multi-Event Online Polarization](https://md.research.elamur.ai/free/en/papers/naseem-2026-polar-benchmark-multilingual-multicultural-multi/)
7. Jonas Pfeiffer, Aishwarya Kamath, Andreas Rücklé, Kyunghyun Cho, and Iryna Gurevych. 2021. [Adapterfusion: Non-destructive task composition for transfer learning](https://md.research.elamur.ai/free/en/papers/pfeiffer-2020-adapterfusion-non-destructive-task-composition/). In Proceedings of the 16th conference of the European chapter of the association for computational linguistics: main volume, pages 487–503…
8. Manoel Horta Ribeiro, Raphael Ottoni, Robert West, Virgílio AF Almeida, and Wagner Meira Jr. 2020. Auditing radicalization pathways on youtube. In Proceedings of the 2020 conference on fairness, accountability, and transparency, pages 131–141. [find](https://scholar.google.com/scholar?q=Manoel+Horta+Ribeiro%2C+Raphael+Ottoni%2C+Robert+West%2C+Virg%C3%ADlio+AF+Almeida%2C+and+Wagner+Meira+Jr%2E+2020%2E+Auditing+radicalization+pathways+on+youtube%2E+In+Proceedings+of+the+2020+conference+on+fairness%2C+accou)
9. Teun A Van Dijk. 1998. Ideology: A multidisciplinary approach. SAGE Publications Ltd. [find](https://scholar.google.com/scholar?q=Teun+A+Van+Dijk%2E+1998%2E+Ideology%3A+A+multidisciplinary+approach%2E+SAGE+Publications+Ltd%2E)
10. Bertie Vidgen and Leon Derczynski. 2020. Directions in abusive language training data, a systematic review: Garbage in, garbage out. Plos one, 15(12):e0243300. [find](https://scholar.google.com/scholar?q=Bertie+Vidgen+and+Leon+Derczynski%2E+2020%2E+Directions+in+abusive+language+training+data%2C+a+systematic+review%3A+Garbage+in%2C+garbage+out%2E+Plos+one%2C+15%2812%29%3Ae0243300%2E)
11. Bertie Vidgen, Tristan Thrush, Zeerak Talat, and Douwe Kiela. 2021. [Learning from the worst: Dynamically generated datasets to improve online hate detection](https://md.research.elamur.ai/free/en/papers/vidgen-2021-learning-worst-dynamically-generated-datasets/). In Proceedings of the 59th annual meeting of the Association for Computational Linguistics and the 11th international joint conference on natu…
12. Ruth Wodak. 2015. The politics of fear: What right-wing populist discourses mean. Sage. [find](https://scholar.google.com/scholar?q=Ruth+Wodak%2E+2015%2E+The+politics+of+fear%3A+What+right%2Dwing+populist+discourses+mean%2E+Sage%2E)

*Updated 31 October 2026*