New framework uses LLMs to boost cross-lingual stance detection for low-resource languages

Researchers have developed a new framework for cross-lingual stance detection that leverages large language models (LLMs) to improve performance in low-resource languages like Catalan. The method uses Chain-of-Thought prompting to guide LLMs in generating reasoning rationales, which are then distilled into a smaller, more efficient student model. This approach aims to overcome the computational costs and latency associated with using LLMs directly, while enhancing stance inference accuracy through a dual-path distillation mechanism and contrastive learning strategies. AI

IMPACT This research could enable more accurate stance detection in languages with limited data, improving downstream applications.

RANK_REASON The cluster contains an academic paper detailing a new methodology for stance detection. [lever_c_demoted from research: ic=1 ai=1.0]

Read on arXiv cs.CL →

• arXiv
• Catalan
• Hugging Face
• large-language models

• paper
• model release

AI-generated summary · Google Gemini · from 1 sources. How we write summaries →