The paper evaluates large language models for hate‑speech detection in Roman Urdu, a low‑resource language with informal spelling variations. Using the Parameter‑Efficient Fine‑Tuning technique Low‑Rank Adaptation (LoRA), the authors fine‑tune models such as Mistral, LLaMA, Falcon, and multilingual BERT on the 72,000‑comment PURUTT dataset. While zero‑shot performance yields an F1 of 0.56, fine‑tuning a small fraction of parameters boosts F1 scores above 0.93, demonstrating that PEFT offers both high accuracy and computational efficiency for low‑resource language tasks.
By Toneema Zubair, Muhammad Junaid Asif, Faisal Kamiran, Hafiz Hassan Saeed, Rana Fayyaz Ahmad
The paper "Ghaib in Translation" investigates how large language models (LLMs) handle Urdu, a widely spoken language that is largely absent from safety evaluations. Five prominent LLMs—GPT‑4o, Claude Sonnet 4.5, Gemini 2.5 Flash, Qwen‑2.5, and Llama‑3.1—were tested on six datasets covering Nastaliq Urdu, Roman Urdu, English, and code‑switched Urdu‑English. The study found significant label instability between original‑script and English‑translation classifications, with missed‑in‑Urdu rates ranging from 2.4% to 9.9% (median 4.3%). A review of 205 papers across nine ALW/WOAH editions revealed no dedicated Urdu research, underscoring the language’s neglect in current safety research.
By Fawzia Zehra (Fuzzy), Kara-Isitt, Sonal Khosla, Stephen Swift
arXiv:2606. 07167v1 Announce Type: cross Abstract: Meaningful multilingual evaluation must test models in the target language and educational context.
By Ahmer Tabassum, Sarfraz Ahmad, Hasan Iqbal, Owais Aijaz, Momina Ahsan, Preslav Nakov
Multilingual Language Models like mBERT are widely used for low-resource NLP, yet their adaptation to morphologically inconsistent languages such as Roman Urdu remains underexplored. Roman Urdu spelling variation causes severe sub-word fragmentation, averaging 1.
Large language models (LLMs) demonstrate impressive performance across a wide range of general NLP tasks; however, their effectiveness in sensitive domains, such as hate speech detection, remains less...
The paper presents an instruction‑tuned large language model (LLM) based on Qwen3 that is fine‑tuned for hate speech mitigation by unifying 36 English hate speech datasets. The authors show that this generalist LLM achieves state‑of‑the‑art performance on in‑domain benchmarks and delivers significant gains in cross‑domain and cross‑lingual generalization, outperforming specialist encoder‑based classifiers.
By Lukas Edman, Daryna Dementieva, Alexander Fraser
arXiv:2609.16393v1 Announce Type: new
Abstract: We introduce ParsHate, a manually annotated dataset of 10,000 Persian tweets spanning 2013-2022, representing the first decade-long benchmark for hate...
By Zahra Bokaei, Walid Magdy, Bonnie Webber
The paper proposes a training‑time explainability framework that aligns model reasoning with human‑annotated rationales to improve both classification performance and interpretability for multilingual hate speech detection. It is evaluated on HateXplain (English) and BullySent (Hinglish), datasets that capture anti‑Muslim hate in culturally coded, multilingual forms. Using methods such as LIME, Integrated Gradients, Grad‑X‑Input, and attention, the study shows that gradient‑ and attention‑based regularization boosts F‑scores, enhances plausibility and faithfulness, and captures culturally specific cues for detecting implicit anti‑Muslim hate.
By Muhammad Deedahwar Mazhar Qureshi, Sannaan Khan, Muhammad Atif Qureshi, Wael Rashwan
The paper evaluates six structured decision models for hate‑speech moderation across four datasets, comparing them to specialized, zero‑shot, commercial, and supervised baselines. It investigates whether providing dataset definitions or decomposing them into multiple questions improves classification, and measures latency and cost. Results show that commercial LLMs outperform decision models on only one dataset, definitions alter up to 28% of predictions without consistent gains, and decomposition helps in only 20% of cases, yet the best hosted decision model approaches the best commercial LLM’s performance with roughly 97% lower inference cost.
By Demetris Paschalides, George Pallis, Marios D. Dikaiakos
The paper introduces BanglaSafe, a benchmark of 879 Bengali prompts that covers 17 culturally grounded harm categories and five prompting conditions. Evaluation of 18 frontier LLMs shows that 53.6% of responses are unsafe or partially unsafe, with 14.7% containing strictly harmful content. The study finds that the writing style within Bengali has a stronger impact on safety than the language switch itself, and that current safety classifiers struggle to reliably evaluate Bengali content.
By Naymul Islam, Nusrat Jahan Lia, Shubhashis Roy Dipta, Sabik Bin Sultan, Abdullah Khan Zehady
We introduce ParsHate, a manually annotated dataset of 10,000 Persian tweets spanning 2013-2022, representing the first decade-long benchmark for hate speech detection in Persian. The dataset contains...
arXiv:2607. 17164v1 Announce Type: new Abstract: Developing Automatic Speech Recognition (ASR) for morphologically rich, low-resource languages such as Assamese is challenging due to insufficient annotated speech data.
By Ganapati Das, Dwipen Laskar, Hasin Afzal Ahmed, Sanjib Kr Kalita, Kshirod Sarmah, Hem Chandra Das, Manjula Kalita