Elamur Research⌕
☰
home / papers / 2026 / RspectNLP@LT-EDI 2026:Rubric-Driven Prompting for Safe Multilingual Counter Narrative Generation

RspectNLP@LT-EDI 2026:Rubric-Driven Prompting for Safe Multilingual Counter Narrative Generation

Paper · 2026 · S.B.Priya and Bharathi B

Summary

The paper addresses Task 2 of the LT-EDI 2026 shared task. The goal is to write constructive, empathetic replies (counter-narratives) to homophobic or transphobic social media comments in English and Tamil, rather than only detecting or deleting such content. The authors build a lightweight zero-shot pipeline around the pretrained google/flan-t5-base model, with no task-specific training. Each harmful comment is placed inside an instruction whose wording mirrors the evaluation rubric. The instruction asks for politeness, relevance and no toxic language, and replies are decoded with beam search. Test items come without labels, so a simple keyword heuristic tags each comment as transphobia or homophobia to fit the submission format. In English the system reaches an overall score of 70.33 with strong contextual coherence. In Tamil it drops to 33.57, with very weak coherence and quality. The study shows that rubric-shaped prompting can cheaply produce safe, on-topic English responses. It also shows how poorly zero-shot generation carries over to a low-resource language like Tamil.

Key points

  • Data: Tamil has 800 training comments (342 homophobic, 458 transphobic) and 109 test comments. English has 1800 training and 66 test comments.
  • Method: google/flan-t5-base is used zero-shot with a rubric-aligned prompt template and beam search decoding. Labels come from keyword rules, and the span field is set to NA.
  • English results: overall 70.33, CCNC (contextual coherence) 81.82 and Distinct-2 78.56. Politeness and quality scores are only moderate.
  • Tamil results: overall 33.57, QS 11.47 and CCNC 7.80. BERTScore-F1 stays comparatively high at 80.23.
  • Evaluation and limitations: scoring combines reference-based metrics (Distinct-2, BERTScore) with rubric-based scores for politeness, quality and coherence. Against a generic prompt, the rubric prompt was reported only qualitatively as better. Proposed future work includes fine-tuning, retrieval augmentation and culturally adapted prompts.

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

 Fields, topics and concepts

📚 Shelves 4The field of science this work belongs to (OpenAlex fields).Computer Science · 16,667 worksArtificial Intelligence · 16,194 worksArts and Humanities · 16,039 worksLanguage and Linguistics · 15,891 works

🧭 Topics 2Research subjects and languages the work is about.Language models & pretraining · 3,320 worksMultilingual & cross-lingual NLP · 828 works

Cite

Cite as S.B.Priya and Bharathi B. 2026. RspectNLP@LT-EDI 2026:Rubric-Driven Prompting for Safe Multilingual Counter Narrative Generation. Proceedings of the Sixth Workshop on Language Technology for Equality, Diversity, Inclusion.

The paper itself, not this page. BibTeX from the library's record of the work. DOI 10.18653/v1/2026.ltedi-1.25

@inproceedings{sbpriya2026rspectnlp,
  title = {{RspectNLP@LT-EDI 2026:Rubric-Driven Prompting for Safe Multilingual Counter Narrative Generation}},
  author = {S.B.Priya and Bharathi B},
  booktitle = {Proceedings of the Sixth Workshop on Language Technology for Equality, Diversity, Inclusion},
  year = {2026},
  doi = {10.18653/v1/2026.ltedi-1.25},
  url = {https://doi.org/10.18653/v1/2026.ltedi-1.25}
}

References

14 references, 0 of them with a card of ours. ↗ leads to the work, «find» searches for one we have no id for yet.

  1. Aish Albladi, Minarul Islam, Amit Das, Maryam Bigo-nah, Zheng Zhang, Fatemeh Jamshidi, Mostafa Rahgouy, Nilanjana Raychawdhary, Daniela Marghitu, and Cheryl Seals. 2025. Hate speech detection using large language models: A comprehensive review. IEEE Access, 13:20871–20892. find
  2. Bharathi Raja Chakravarthi. 2024. Detection of homophobia and transphobia in youtube comments. In-https://github.com/Prisur2013/RESPECTNLP_1991 ternational Journal of Data Science and Analytics, 18(1):49–68. find
  3. Koyel Ghosh, Saptarshi Saha, Thomas Mandl, and Sandip Modha. 2025. Findings from shared tasks on hate speech detection: Performance patterns for low-resource languages. Pattern Recognition Letters. find
  4. Koyel Ghosh and Apurbalal Senapati. 2025. Hate speech detection in low-resourced indian languages: An analysis of transformer-based monolingual and multilingual models with cross-lingual experiments. Natural language processing, 31(2):393–414. find
  5. Shuyu Jiang, Wenyi Tang, Xingshu Chen, Rui Tang, Haizhou Wang, and Wenxian Wang. 2025. Rezg: Retrieval-augmented zero-shot counter narrative generation for hate speech. Neurocomputing, 620:129140. vulnerable populations using llms with enhanced safety, fairness, and controllability. arXiv preprint a… — [arXiv:2509.08910](https://arxiv.org/abs/2509.08910) find
  6. Sahil Wadhwa, Chengtian Xu, Haoming Chen, Aakash Mahalingam, Akankshya Kar, and Divya Chaudhary. 2025. Northeastern uni at multilingual counterspeech generation: Enhancing counter speech generation with llm alignment through direct preference optimization. In Proceedings of the First Workshop on Mul… find
  7. Brian Wilk, Homaira Huda Shomee, Suman Kalyan Maity, and Sourav Medya. 2025. Fact-based counter narrative generation to combat hate speech. In Proceedings of the ACM on Web Conference 2025, pages 3354–3365. find
  8. Prasanna Kumar Kumaresan, Devendra Deepak Kayande, Ruba Priyadharshini, Paul Buitelaar, and Bharathi Raja Chakravarthi. 2025. Homophobia and transphobia span identification in low-resource languages. Natural Language Processing Journal, page 100169. find
  9. Prasanna Kumar Kumaresan, Praveen Prasannan, Tanay Singh, Ruba Priyadharshini, Subalalitha Chinnaudayar Navaneethakrishnan, Saranya Rajiakodi, Paul Buitelaar, and Bharathi Raja Chakravarthi. 2026. Findings of the Shared Task on Counter-Narrative Generation on Homophobic and Transphobic Comments. In… find
  10. Khouloud Mnassri, Reza Farahbakhsh, and Noel Crespi. 2024. Multilingual hate speech detection: a semi-supervised generative adversarial approach. Entropy, 26(4):344. find
  11. Praveen Prasannan, Prasanna Kumar Kumaresan, Saranya Rajiakodi, CN Subalalitha, and Bharathi Raja Chakravarthi. 2025. Counter-speech generation for homophobic and transphobic social media content in malayalam. Social Network Analysis and Mining, 15(1):87. find
  12. Hafiz Muhammad Raza Ur Rehman, Mahpara Saleem, Muhammad Zeeshan Jhandir, Eduardo Silva Alvarado, Helena Garay, and Imran Ashraf. 2025. Detecting hate in diversity: a survey of multilingual code-mixed image and video analysis. Journal of Big Data, 12(1):109. find
  13. Furqan Khan Saddozai, Sahar K Badri, Daniyal Alghaz-zawi, Asad Khattak, and Muhammad Zubair Asghar. 2025. Multimodal hate speech detection: a novel deep learning framework for multilingual text and images. PeerJ Computer Science, 11:e2801. find
  14. Tung Vu, Lam Nguyen, and Quynh Dao. 2025. Promptguard: An orchestrated prompting framework for principled synthetic text generation for find

Updated

↑ back to top

This page in Markdown → https://md.research.elamur.ai/free/en/papers/priya-2026-rspectnlp-lt-edi-2026-rubric/