arXiv:2511.01265v3 Announce Type: replace 
Abstract: We introduce AraFinNews, the largest publicly available Arabic financial news dataset to date, comprising 212,500 article-headline pairs spanning a decade of reporting from 2015 to 2025. Designed as an Arabic counterpart to major English summarisation corpora such as CNN/DailyMail, AraFinNews provides a realistic benchmark for evaluating domain-specific language understanding and generation in financial contexts. Using this resource, we investigate the impact of domain specificity on abstractive summarisation of Arabic financial texts with large language models (LLMs). In particular, we evaluate transformer-based models: mT5, AraT5, and the domain-adapted FinAraT5 to examine how financial-domain pretraining influences accuracy, numerical reliability, and stylistic alignment with professional reporting. Experimental results show that domain-adapted models generate more coherent summaries, especially in their handling of quantitative and entity-centric information. These findings highlight the importance of domain-specific adaptation for improving narrative fluency in Arabic financial summarisation. The dataset is freely available for non-commercial research at https://github.com/ArabicNLP-uk/AraFinNews.

تم تقديم AraFinNews كأكبر مجموعة بيانات للأخبار المالية العربية متاحة للجمهور، حيث تضم 212,500 زوج من المقالات والعناوين من عام 2015 إلى 2025، بهدف تحسين تلخيص النصوص المالية العربية باستخدام نماذج اللغة الكبيرة (LLMs). وتعتبر هذه المجموعة مرجعًا لتقييم فهم اللغة وتوليدها في السياقات المالية، خاصة من خلال النماذج المعتمدة على المحولات مثل mT5 وAraT5 وFinAraT5.

AraFinNews se ha presentado como el conjunto de datos de noticias financieras árabes más grande disponible públicamente, que incluye 212,500 pares de artículo-título desde 2015 hasta 2025, con el objetivo de mejorar la resumación de textos financieros árabes utilizando modelos de lenguaje de gran tamaño (LLMs). Este conjunto de datos sirve como un referente para evaluar la comprensión y generación de lenguaje en contextos financieros, especialmente a través de modelos basados en transformadores como mT5, AraT5 y FinAraT5.

AraFinNews a été introduit comme le plus grand ensemble de données de nouvelles financières arabes disponible publiquement, comprenant 212 500 paires article-titre de 2015 à 2025, visant à améliorer la résumation de textes financiers arabes à l'aide de modèles de langage de grande taille (LLMs). Cet ensemble de données sert de référence pour évaluer la compréhension et la génération de langage dans des contextes financiers, en particulier à travers des modèles basés sur des transformateurs comme mT5, AraT5 et FinAraT5.

AraFinNews has been introduced as the largest publicly available Arabic financial news dataset, featuring 212,500 article-headline pairs from 2015 to 2025, aimed at enhancing Arabic financial text summarization using large language models (LLMs). The dataset serves as a benchmark for evaluating language understanding and generation in financial contexts, particularly through transformer-based models like mT5, AraT5, and FinAraT5.

AraFinNews: Arabic Financial Summarisation with Domain-Adapted LLMs

arXiv:2508.09179v2 Announce Type: replace-cross 
Abstract: Reconstructing high-fidelity MR images from undersampled k-space data remains a challenging problem in MRI. While Mamba variants for vision tasks offer promising long-range modeling capabilities with linear-time complexity, their direct application to MRI reconstruction inherits two key limitations: (1) insensitivity to high-frequency anatomical details; and (2) reliance on redundant multi-directional scanning. To address these limitations, we introduce High-Fidelity Mamba (HiFi-Mamba), a novel dual-stream Mamba-based architecture comprising stacked W-Laplacian (WL) and HiFi-Mamba blocks. Specifically, the WL block performs fidelity-preserving spectral decoupling, producing complementary low- and high-frequency streams. This separation enables the HiFi-Mamba block to focus on low-frequency structures, enhancing global feature modeling. Concurrently, the HiFi-Mamba block selectively integrates high-frequency features through adaptive state-space modulation, preserving comprehensive spectral details. To eliminate the scanning redundancy, the HiFi-Mamba block adopts a streamlined unidirectional traversal strategy that preserves long-range modeling capability with improved computational efficiency. Extensive experiments on standard MRI reconstruction benchmarks demonstrate that HiFi-Mamba consistently outperforms state-of-the-art CNN-based, Transformer-based, and other Mamba-based models in reconstruction accuracy while maintaining a compact and efficient model design.

تمثل مقدمة HiFi-Mamba، وهي بنية معمارية تعتمد على Mamba ذات تدفق مزدوج، خطوة نحو تحسين إعادة بناء صور الرنين المغناطيسي عالية الدقة من بيانات k-space غير المأخوذة بعين الاعتبار، من خلال معالجة القيود الرئيسية للمتغيرات الحالية من Mamba. تتضمن البنية كتل W-Laplacian وHiFi-Mamba المكدسة، التي تفصل بين تدفقات الترددات المنخفضة والعالية لتحسين دقة الصورة وتفاصيلها.

La introducción de HiFi-Mamba, una arquitectura basada en Mamba de doble flujo, tiene como objetivo mejorar la reconstrucción de MRI de alta fidelidad a partir de datos de k-space submuestreados, abordando limitaciones clave de las variantes de Mamba existentes. La arquitectura presenta bloques apilados de W-Laplaciano y HiFi-Mamba, que separan los flujos de baja y alta frecuencia para mejorar la fidelidad y el detalle de las imágenes.

L'introduction de HiFi-Mamba, une architecture basée sur Mamba à double flux, vise à améliorer la reconstruction d'IRM haute fidélité à partir de données k-space sous-échantillonnées en abordant les limitations clés des variantes Mamba existantes. L'architecture comprend des blocs W-Laplacien et HiFi-Mamba empilés, qui séparent les flux de basses et hautes fréquences pour améliorer la fidélité et le détail des images.

The introduction of HiFi-Mamba, a dual-stream Mamba-based architecture, aims to enhance high-fidelity MRI reconstruction from undersampled k-space data by addressing key limitations of existing Mamba variants. The architecture features stacked W-Laplacian and HiFi-Mamba blocks, which separate low- and high-frequency streams to improve image fidelity and detail.

AraFinNews: Arabic Financial Summarisation with Domain-Adapted LLMs

Was this article worth reading? Share it

finlight.me

LucidQuery AI

Airparser

Artificial intelligence

Syft AI

NewsperAI

Ready to build your own newsroom?