Researchers have developed Riazi-8B, a new large language model specifically designed for mathematical reasoning in Urdu. This model was created through a two-step adaptation process, involving continued pre-training on Urdu Wikipedia and fine-tuning on Urdu Chain-of-Thought data. Evaluations on the MGSM-Urdu benchmark show that Riazi-8B outperforms existing Urdu instruction-tuned models in answer correctness and reasoning quality, demonstrating an effective strategy for extending advanced AI capabilities to low-resource languages. AI
IMPACT Extends advanced mathematical reasoning capabilities of LLMs to low-resource languages like Urdu.
RANK_REASON The cluster describes the release of a new research paper detailing a specialized LLM for a low-resource language. [lever_c_demoted from research: ic=1 ai=1.0]
AI-generated summary · Google Gemini · from 2 sources. How we write summaries →
