Natural language processing research from GCC institutions, covering Arabic NLP, multilingual models, text classification, named entity recognition, and machine translation.
The paper introduces AraPoemBERT, an Arabic language model pretrained exclusively on 2.09 million verses of Arabic poetry. AraPoemBERT was evaluated against five other Arabic language models on tasks including poet's gender classification (99.34% accuracy) and poetry sub-meter classification (97.79% accuracy). The model achieved state-of-the-art results in these and other downstream tasks, and is publicly available on Hugging Face. Why it matters: This specialized model advances Arabic NLP by providing a new state-of-the-art tool tailored for the nuances of classical Arabic poetry.
This paper explores the impact of tokenization strategies and vocabulary sizes on Arabic language model performance across NLP tasks like news classification and sentiment analysis. It compares four tokenizers, finding that Byte Pair Encoding (BPE) with Farasa performs best overall due to its morphological analysis capabilities. The study surprisingly found limited impact of vocabulary size on performance with fixed model sizes, challenging assumptions about vocabulary size and model performance. Why it matters: The findings provide insights for developing more effective and nuanced Arabic language models, particularly for handling dialectal variations and promoting responsible AI development in the region.
The paper introduces AraTrust, a new benchmark for evaluating the trustworthiness of LLMs when prompted in Arabic. The benchmark contains 522 multiple-choice questions covering dimensions like truthfulness, ethics, safety, and fairness. Experiments using AraTrust showed that GPT-4 performed the best, while open-source models like AceGPT 7B and Jais 13B had lower scores. Why it matters: This benchmark addresses a critical gap in evaluating LLMs for Arabic, which is essential for ensuring the safe and ethical deployment of AI in the Arab world.
Researchers from MBZUAI have released MobiLlama, a fully transparent open-source 0.5 billion parameter Small Language Model (SLM). MobiLlama is designed for resource-constrained devices, emphasizing enhanced performance with reduced resource demands. The full training data pipeline, code, model weights, and checkpoints are available on Github.
The paper introduces ArabianGPT, a suite of transformer-based language models designed specifically for Arabic, including versions with 0.1B and 0.3B parameters. A key component is the AraNizer tokenizer, tailored for Arabic script's morphology. Fine-tuning ArabianGPT-0.1B achieved 95% accuracy in sentiment analysis, up from 56% in the base model, and improved F1 scores in summarization. Why it matters: The models address the gap in native Arabic LLMs, offering better performance on Arabic NLP tasks through tailored architecture and tokenization.
Researchers introduce PALO, a polyglot large multimodal model with visual reasoning capabilities in 10 major languages including Arabic. A semi-automated translation approach was used to adapt the multimodal instruction dataset from English to the target languages. The models are trained across three scales (1.7B, 7B and 13B parameters) and a multilingual multimodal benchmark is proposed for evaluation.
MBZUAI researchers introduce BiMediX, a bilingual (English and Arabic) mixture of experts LLM for medical applications. The model is trained on BiMed1.3M, a new 1.3 million bilingual instruction dataset and outperforms existing models like Med42 and Jais-30B on medical benchmarks. Code and models are available on Github.
This paper investigates the intrinsic self-correction capabilities of LLMs, identifying model confidence as a key latent factor. Researchers developed an "If-or-Else" (IoE) prompting framework to guide LLMs in assessing their own confidence and improving self-correction accuracy. Experiments demonstrate that the IoE-based prompt enhances the accuracy of self-corrected responses, with code available on GitHub.
MBZUAI researchers introduce M4GT-Bench, a new benchmark for evaluating machine-generated text (MGT) detection across multiple languages and domains. The benchmark includes tasks for binary MGT detection, identifying the specific model that generated the text, and detecting mixed human-machine text. Experiments with baseline models and human evaluation show that MGT detection performance is highly dependent on access to training data from the same domain and generators.
Researchers from the National Center for AI in Saudi Arabia investigated the sensitivity of Large Language Model (LLM) leaderboards to minor benchmark perturbations. They found that small changes, like choice order, can shift rankings by up to 8 positions. The study recommends hybrid scoring and warns against over-reliance on simple benchmark evaluations, providing code for further research.
This paper studies the impact of data scale on Arabic Pretrained Language Models (PLMs). Researchers retrained BERT-base and T5-base models on large Arabic corpora, achieving state-of-the-art results on the ALUE and ORCA benchmarks. The analysis indicates that pretraining data volume is the most important factor for performance. Why it matters: This work provides valuable insights into building effective Arabic language models, emphasizing the importance of large, high-quality datasets for advancing Arabic NLP.
Researchers introduce Arabic Mini-ClimateGPT, a tailored Arabic LLM for climate change and sustainability. The model is fine-tuned on the Clima500-Instruct dataset and uses vector embedding retrieval during inference. Evaluations show the model outperforms baseline LLMs and is preferred by experts in 81.6% of cases.
KAUST researchers found Y-series nonfullerene acceptors enhance the outdoor stability of organic solar cells, enabling energy-efficient windows. They also used satellite data to show managed vegetation can mitigate rising temperatures across Saudi Arabia's agricultural regions. Additionally, they developed DeepKriging, a deep neural network, to solve complex spatiotemporal datasets and tested it on air pollution. Why it matters: This research addresses critical challenges in renewable energy, climate change, and AI data privacy relevant to Saudi Arabia and the broader region.
This paper introduces a predictive analysis of Arabic court decisions, utilizing 10,813 real commercial court cases. The study evaluates LLaMA-7b, JAIS-13b, and GPT3.5-turbo models under zero-shot, one-shot, and fine-tuned training paradigms, also experimenting with summarization and translation. GPT-3.5 models significantly outperformed others, exceeding JAIS model performance by 50%, while also demonstrating the unreliability of most automated metrics. Why it matters: This research bridges computational linguistics and Arabic legal analytics, offering insights for enhancing judicial processes and legal strategies in the Arabic-speaking world.
Researchers introduce AceGPT, a localized large language model (LLM) specifically for Arabic, addressing cultural sensitivity and local values not well-represented in mainstream models. AceGPT incorporates further pre-training with Arabic texts, supervised fine-tuning using native Arabic instructions and GPT-4 responses, and reinforcement learning with AI feedback using a reward model attuned to local culture. Evaluations demonstrate that AceGPT achieves state-of-the-art performance among open Arabic LLMs across several benchmarks. Why it matters: This work advances culturally-aware AI development for Arabic-speaking communities, providing a valuable resource and benchmark for future research.
Researchers at MBZUAI release SlimPajama-DC, an empirical analysis of data combinations for pretraining LLMs using the SlimPajama dataset. The study examines the impact of global vs. local deduplication and the proportions of highly-deduplicated multi-source datasets. Results show that increased data diversity after global deduplication is crucial, with the best configuration outperforming models trained on RedPajama.
G42's Core42 has released Jais, a collection of Arabic large language models, including a 13B parameter version. Jais-13B is trained on a 395B token dataset containing Arabic and English text. According to the blog post, Jais-13B achieves state-of-the-art results on Arabic NLP benchmarks. Why it matters: This release establishes a new benchmark for Arabic language AI, potentially enabling more sophisticated and culturally relevant applications.
G42's Core42 has released Jais, a new Arabic large language model. Jais includes 13 billion parameters and was trained on a dataset of 126B tokens, including 43B Arabic tokens. According to the developers, Jais achieves state-of-the-art results on Arabic benchmarks and competitive performance on English benchmarks. Why it matters: Jais represents a significant step forward for Arabic NLP, providing a powerful new tool for researchers and developers in the region.
MBZUAI has released Jais and Jais-chat, two new open generative large language models (LLMs) with a focus on Arabic. The 13 billion parameter models are based on the GPT-3 architecture and pretrained on Arabic, English, and code. Evaluation shows state-of-the-art Arabic knowledge and reasoning, with competitive English performance.
MBZUAI researchers introduce XrayGPT, a conversational medical vision-language model for analyzing chest radiographs and answering open-ended questions. The model aligns a medical visual encoder (MedClip) with a fine-tuned large language model (Vicuna) using a linear transformation. To enhance performance, the LLM was fine-tuned using 217k interactive summaries generated from radiology reports.
Video-ChatGPT is a new multimodal model that combines a video-adapted visual encoder with a large language model (LLM) to enable detailed video understanding and conversation. The authors introduce a new dataset of 100,000 video-instruction pairs for training the model. They also develop a quantitative evaluation framework for video-based dialogue models.
The paper introduces Sparse-Quantized Representation (SpQR), a new compression format and quantization technique for large language models (LLMs). SpQR identifies outlier weights and stores them in higher precision while compressing the remaining weights to 3-4 bits. The method achieves less than 1% accuracy loss in perplexity for LLaMA and Falcon LLMs and enables a 33B parameter LLM to run on a single 24GB consumer GPU. Why it matters: This enables near-lossless compression of LLMs, making powerful models accessible on resource-constrained devices and accelerating inference without significant accuracy degradation.
This paper presents a comprehensive evaluation of ChatGPT's performance across 44 Arabic NLP tasks using over 60 datasets. The study compares ChatGPT's capabilities in Modern Standard Arabic (MSA) and Dialectal Arabic (DA) against smaller, fine-tuned models. Results show ChatGPT is outperformed by smaller, fine-tuned models and exhibits limitations in handling Arabic dialects compared to MSA. Why it matters: The work highlights the need for further research and development of Arabic-specific NLP models to overcome the limitations of general-purpose models like ChatGPT.
LAraBench introduces a benchmark for Arabic NLP and speech processing, evaluating LLMs like GPT-3.5-turbo, GPT-4, BLOOMZ, Jais-13b-chat, Whisper, and USM. The benchmark covers 33 tasks across 61 datasets, using zero-shot and few-shot learning techniques. Results show that SOTA models generally outperform LLMs in zero-shot settings, though larger LLMs with few-shot learning reduce the gap. Why it matters: This benchmark helps assess and improve the performance of LLMs on Arabic language tasks, highlighting areas where specialized models still excel.
MBZUAI researchers introduce M4, a multi-generator, multi-domain, and multi-lingual benchmark dataset for detecting machine-generated text. The study reveals challenges in generalizing detection across unseen domains or LLMs, with detectors often misclassifying machine-generated text as human-written. The dataset aims to foster research into more robust detection methods and is available on GitHub.
MBZUAI releases Bactrian-X, a multilingual parallel dataset of 3.4 million instruction-response pairs across 52 languages. They trained low-rank adaptation (LoRA) adapters using this dataset, creating lightweight, replaceable components for large language models. Experiments show the LoRA-based models outperform vanilla and existing instruction-tuned models in multilingual settings.
This paper introduces DetectLLM-LRR and DetectLLM-NPR, two novel zero-shot methods for detecting machine-generated text using log rank information. Experiments across three datasets and seven language models demonstrate improvements of up to 3.9 AUROC points over state-of-the-art methods. The code and data for both methods are available on Github.
This paper introduces ProgramFC, a fact-checking model that decomposes complex claims into simpler sub-tasks using a library of functions. The model uses LLMs to generate reasoning programs and executes them by delegating sub-tasks, enhancing explainability and data efficiency. Experiments on fact-checking datasets demonstrate ProgramFC's superior performance compared to baseline methods, with publicly available code and data.
This paper introduces a Regulatory Knowledge Graph (RKG) for the Abu Dhabi Global Market (ADGM) regulations, constructed using language models and graph technologies. A portion of the regulations was manually tagged to train BERT-based models, which were then applied to the rest of the corpus. The resulting knowledge graph, stored in Neo4j, and code are open-sourced on GitHub to promote advancements in compliance automation.
This paper introduces Provable Unrestricted Adversarial Training (PUAT), a novel adversarial training approach. PUAT enhances robustness against both unrestricted and restricted adversarial examples while improving standard generalizability by aligning the distributions of adversarial examples, natural data, and the classifier's learned distribution. The approach uses partially labeled data and an augmented triple-GAN to generate effective unrestricted adversarial examples, demonstrating superior performance on benchmarks.
The paper introduces ORCA, a new public benchmark for evaluating Arabic language understanding. ORCA covers diverse Arabic varieties and includes 60 datasets across seven NLU task clusters. The benchmark was used to compare 18 multilingual and Arabic language models and includes a public leaderboard with a unified evaluation metric. Why it matters: ORCA addresses the lack of a comprehensive Arabic benchmark, enabling better progress measurement for Arabic and multilingual language models.
The paper introduces LLMEffiChecker, a tool to test the computational efficiency robustness of LLMs by identifying vulnerabilities that can significantly degrade performance. LLMEffiChecker uses both white-box (gradient-guided perturbation) and black-box (causal inference-based perturbation) methods to delay the generation of the end-of-sequence token. Experiments on nine public LLMs demonstrate that LLMEffiChecker can substantially increase response latency and energy consumption with minimal input perturbations.
Qatar Computing Research Institute (QCRI) has developed NatiQ, an end-to-end text-to-speech (TTS) system for Arabic utilizing encoder-decoder architectures. The system employs Tacotron-based models and Transformer models to generate mel-spectrograms, which are then synthesized into waveforms using vocoders like WaveRNN, WaveGlow, and Parallel WaveGAN. Trained on in-house speech data featuring a neutral male voice (Hamza) and an expressive female voice (Amina), NatiQ achieves a Mean Opinion Score (MOS) of 4.21 and 4.40, respectively. Why it matters: This research advances Arabic language technology, providing high-quality TTS synthesis that can enhance accessibility and usability of digital content for Arabic speakers.
This paper introduces a method for automatically designing convolutional neural network (CNN) architectures tailored for diabetic retinopathy (DR) diagnosis using fundus images. The approach uses k-medoid clustering, PCA, and inter/intra-class variations to optimize CNN depth and width. Validated on datasets including a local Saudi dataset and Kaggle benchmarks, the custom-designed models outperform pre-trained CNNs with fewer parameters.
This paper introduces a unified deep autoregressive model (UAE) for cardinality estimation that learns joint data distributions from both data and query workloads. It uses differentiable progressive sampling with the Gumbel-Softmax trick to incorporate supervised query information into the deep autoregressive model. Experiments show UAE achieves better accuracy and efficiency compared to state-of-the-art methods.
The Qatar Computing Research Institute (QCRI) has released QASR, a 2,000-hour transcribed Arabic speech corpus collected from Aljazeera news broadcasts. The dataset features multi-dialect speech sampled at 16kHz, aligned with lightly supervised transcriptions and linguistically motivated segmentation. QCRI also released a 130M word dataset to improve language model training. Why it matters: QASR enables new research in Arabic speech recognition, dialect identification, punctuation restoration, and other NLP tasks for spoken data.
The paper introduces AraELECTRA, a new Arabic language representation model. AraELECTRA is pre-trained using the replaced token detection objective on large Arabic text corpora. The model is evaluated on multiple Arabic NLP tasks, including reading comprehension, sentiment analysis, and named-entity recognition. Why it matters: AraELECTRA outperforms current state-of-the-art Arabic language representation models, given the same pretraining data and even with a smaller model size, advancing Arabic NLP.
The paper introduces AraGPT2, a suite of pre-trained transformer models for Arabic language generation, with the largest model (AraGPT2-mega) containing 1.46 billion parameters. Trained on a large Arabic corpus of internet text and news, AraGPT2-mega demonstrates strong performance in synthetic news generation and zero-shot question answering. To address the risk of misuse, the authors also released a discriminator model with 98% accuracy in detecting AI-generated text. Why it matters: This release of both the model and discriminator fills a critical gap in Arabic NLP and encourages further research and applications in the field.
A retrospective study in Abu Dhabi, UAE, developed a machine learning-based prognostic system to predict the risk of seven complications in COVID-19 patients using data from 3,352 patient encounters. The system, trained on data from the first 24 hours of admission, achieved high accuracy (AUROC > 0.80) in predicting complications like AKI, ARDS, and elevated biomarkers in geographically split test sets. The models primarily used gradient boosting and logistic regression.
This paper introduces GigaBERT, a customized bilingual BERT model pre-trained for Arabic NLP and English-to-Arabic zero-shot transfer learning. The study evaluates GigaBERT's performance on four information extraction tasks: named entity recognition, part-of-speech tagging, argument role labeling, and relation extraction. Results show that GigaBERT outperforms mBERT, XLM-RoBERTa, and AraBERT in both supervised and zero-shot transfer settings. Why it matters: GigaBERT advances Arabic NLP by providing a high-performing, publicly available model tailored for the complexities of the Arabic language and cross-lingual applications.
Researchers at the American University of Beirut (AUB) have released AraBERT, a BERT model pre-trained specifically for Arabic language understanding. The model was trained on a large Arabic corpus and compared against multilingual BERT and other state-of-the-art methods. AraBERT achieved state-of-the-art performance on several tested Arabic NLP tasks including sentiment analysis, named entity recognition, and question answering. Why it matters: This release provides the Arabic NLP community with a high-performing, open-source language model, facilitating further research and development.
Researchers introduce AraNet, a deep learning toolkit for Arabic social media processing. The toolkit uses BERT models trained on social media datasets to predict age, dialect, gender, emotion, irony, and sentiment. AraNet achieves state-of-the-art or competitive performance on these tasks without feature engineering. Why it matters: The public release of AraNet accelerates Arabic NLP research by providing a comprehensive, deep learning-based tool for various social media analysis tasks.
This paper introduces a large-scale historical corpus of written Arabic spanning 1400 years. The corpus was cleaned and processed using Arabic NLP tools, including identification of reused text. The study uses a novel automatic periodization algorithm to study the history of the Arabic language, confirming the division into Modern Standard and Classical Arabic. Why it matters: This resource enables further computational research into the evolution of Arabic and the development of NLP tools for historical texts.
CSAIL (MIT's Computer Science and Artificial Intelligence Laboratory) and the Qatar Computing Research Institute (QCRI) have announced a new research collaboration. This partnership aims to foster joint research projects and facilitate knowledge exchange between the two prominent AI research institutions. The collaboration is expected to leverage expertise from both sides in various AI domains. Why it matters: This partnership signifies a strengthening of AI research ties between a leading US institution and a major Middle Eastern research hub, potentially driving advancements relevant to the region and the broader AI field.
A new paper coauthored by researchers at The University of Melbourne and MBZUAI explores disagreement in human annotation for AI training. The paper treats disagreement as a signal (human label variation or HLV) rather than noise, and proposes new evaluation metrics based on fuzzy set theory. These metrics adapt accuracy and F-score to cases where multiple labels may plausibly apply, aligning model output with the distribution of human judgments. Why it matters: This research addresses a key challenge in NLP by accounting for the inherent ambiguity in human language, potentially leading to more robust and human-aligned AI systems.
A new paper from MBZUAI introduces JEEM, a benchmark dataset for evaluating vision-language models on their understanding of images grounded in four Arabic-speaking societies (Jordan, UAE, Egypt, and Morocco) and their ability to use local dialects. The dataset comprises 2,178 images and 10,890 question-answer pairs reflecting everyday life and culturally specific scenes. Evaluation of several Arabic-capable models (Maya, PALO, Peacock, AIN, AyaV) and GPT-4o revealed that while models can generate fluent language, they struggle with genuine understanding, consistency, and relevance, especially when cultural context is important. Why it matters: This research highlights the challenges of building AI systems that can truly understand and interact with diverse cultures, emphasizing the need for culturally grounded datasets and evaluation metrics.
An MBZUAI team led by Thamar Solorio and Monojit Choudhury received a Google Academic Research Award to study how AI can better understand and respond to human loneliness in digital spaces. The project will examine how loneliness is expressed online, how conversational agents can detect it, and what healthier AI companionship could look like in collaboration with Georgia Tech. The team aims to define digital loneliness and its expression in online conversations with AI. Why it matters: This research addresses a growing global issue by exploring the ethical and psychological implications of AI companionship, potentially leading to safer and more beneficial AI interactions.
MBZUAI has published 674 papers in 2023 and holds a global ranking of 18 in AI, CV, ML, NLP, and robotics according to CSRankings. The university presented 30 papers at ICCV, will present 53 papers at NeurIPS, and has 44 papers at EMNLP 2023. MBZUAI was also awarded its first patent by the US Patent Office for a system and method for handwriting generation. Why it matters: This demonstrates the rapid growth and increasing prominence of MBZUAI as a leading AI research institution in the region and globally.
MBZUAI researchers have developed an AI program using vision transformers that can learn a person's handwriting style and generate text in that style. The US Patent and Trademark Office recently granted a patent for this technology, which could aid individuals with writing impairments. The system overcomes limitations of previous GAN-based approaches by processing long-range dependencies in handwriting. Why it matters: This patented AI tool enhances personalized text generation and has potential applications in assistive technology and improving handwriting recognition models.
MBZUAI researchers developed LLMVoX, a system enabling LLMs to produce real-time speech, including Arabic. LLMVoX addresses limitations of existing end-to-end and cascaded pipeline approaches, which suffer from either degraded reasoning or latency. LLMVoX was developed as part of Project OMER, which was recently awarded Regional Research Grant from Meta. Why it matters: This enhances the potential of LLMs to function as more natural, multimodal virtual assistants, especially for Arabic-speaking users in the Middle East.