Topics ›
General AI research publications from MBZUAI, KAUST, TII, QCRI, and other GCC institutions, including conference papers, journal articles, and preprints.
651–700 articles · Page 14 RSS ↗
MBZUAI researchers introduce PG-Video-LLaVA, a large multimodal model with pixel-level grounding capabilities for videos, integrating audio cues for enhanced understanding. The model uses an off-the-shelf tracker and grounding module to localize objects in videos based on user prompts. PG-Video-LLaVA is evaluated on video question-answering and grounding benchmarks, using Vicuna instead of GPT-3.5 for reproducibility.
KAUST · · Infrastructure Research
KAUST's Shaheen III has been confirmed as the Middle East’s most powerful supercomputer, ranking 20th globally. Built by HPE, Shaheen III is six times faster than KAUST’s existing Shaheen II system, exceeding the processing power of 500,000 MacBook Pros. It will support mathematical model construction for scientific discovery, engineering design, and policy support, particularly in sustainability and alignment with Vision 2030. Why it matters: This positions KAUST and Saudi Arabia as leaders in high-performance computing, enabling advanced research across climate change, clean energy, and other critical fields.
KAUST · · Research Healthcare
A KAUST study explains how corals control their symbiotic algae using nutrient competition, tracking carbon and nitrogen isotopes. The research shows that cnidarians limit nitrogen available to the algae, preventing overgrowth and maintaining a balanced symbiosis. This mechanism evolved independently in corals, sea anemones, and jellyfish. Why it matters: The findings have implications for coral reef restoration efforts like the KAUST Reefscape Restoration Initiative by disrupting traditional views of symbiosis.
Researchers at the Technology Innovation Institute (TII) have released a fully-annotated dataset for autonomous drone racing, called "Race Against the Machine." The dataset includes high-resolution visual, inertial, and motion capture data from both autonomous and piloted flights, along with commands, control inputs, and corner-level labeling of drone racing gates. The specifications to recreate their flight platform using commercial off-the-shelf components and the Betaflight controller are also released. Why it matters: This comprehensive resource aims to support the development of new methods and establish quantitative comparisons for approaches in robotics and AI, democratizing drone racing research.
KAUST · · Research Partnership
KAUST and Chinese companies Shandong Lianxin Environmental Protection Technology and Hangzhou Hecai Technology will manufacture green plastics based on KAUST technology. The plastics, high molar mass aliphatic polycarbonates, are for biomedical products and food packaging due to their biodegradability and biocompatibility. KAUST's method creates these polycarbonates using CO2 and sustainable raw materials without toxic metals, with production scaling over two years. Why it matters: This partnership highlights KAUST's role in developing sustainable materials and bringing them to market, with potential impact on reducing reliance on traditional plastics in sensitive applications.
KAUST · · Healthcare Research
Researchers from KAUST, King Faisal Specialist Hospital, and collaborators have developed a new method to predict cardiometabolic disease risk in underrepresented ethnic populations using genetic information and public databases. The study focused on Arab communities and created a framework to determine polygenic scores for more accurate heart disease prediction. The framework was validated using records of over 5,000 Arab patients, demonstrating that genetic risk complements conventional risk factors. Why it matters: This research addresses a critical gap in genomic data for non-European populations, potentially leading to more effective and personalized healthcare strategies in the Arab world and beyond.
KAUST · · Partnership Research
Ceer and KAUST are partnering to research and develop innovations in smart mobility, including connectivity and autonomous driving technologies. The collaboration aims to transform research into economically productive innovations, supporting Saudi-based SMEs and the automotive manufacturing industry. Ceer and KAUST will share facilities and conduct collaborative activities to foster knowledge exchange and create jobs. Why it matters: This partnership signifies a crucial step towards diversifying Saudi Arabia's economy and advancing technological capabilities in the electric vehicle sector within the Kingdom.
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.
KAUST, AEON Collective, and KAPSARC released a report during MENA Climate Week assessing climate change impacts in Saudi Arabia under a 3°C warming scenario. The report finds that Saudi Arabia is experiencing climate change at an accelerated pace, with potential temperature increases of up to 5.6°C by the end of the century. It highlights the need for policy choices and adaptation strategies to mitigate impacts on ecosystems, urban environments, and human health. Why it matters: This comprehensive analysis provides critical insights for Saudi Arabia to inform its climate policies and adaptation efforts, addressing challenges like coral bleaching, desert expansion, and increased health risks.
KAUST · · Research Partnership
KAUST and Cerebras Systems collaborated on multi-dimensional seismic processing using the Condor Galaxy AI supercomputer, achieving record sustained memory bandwidth of 92.58 petabytes per second. They developed a Tile Low-Rank Matrix-Vector Multiplication (TLR-MVM) kernel to exploit the architecture of Cerebras CS-2 systems. This work was recognized as a finalist for the 2023 Gordon Bell Prize. Why it matters: This demonstrates the potential of AI-customized architectures for seismic processing, with broader implications for climate modeling and other scientific domains in the region and globally.
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.
This paper introduces a domain generalization (DG) method for Diabetic Retinopathy (DR) classification that maximizes mutual information using a large pretrained model. The method aims to address the challenge of domain shift in medical imaging caused by variations in data acquisition. Experiments on public datasets demonstrate that the proposed method outperforms state-of-the-art techniques, achieving a 5.25% improvement in average accuracy.
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.
arXiv · · Healthcare CV
This paper introduces Pulmonary Embolism Detection using Contrastive Learning (PECon), a supervised contrastive pretraining strategy using both CT scans and EHR data to improve feature alignment between modalities for better PE diagnosis. PECon pulls sample features of the same class together while pushing away features of other classes. The approach achieves state-of-the-art results on the RadFusion dataset, with an F1-score of 0.913 and AUROC of 0.943.
Excyton, a startup based at KAUST, has developed a novel display technology called “TurboLED” that reduces power consumption by 50% and increases the color range rendered on displays to 76%. The technology utilizes a six sub-pixel format (light and deep RGB) compared to the standard three, saving energy by using lighter colors most of the time. Excyton received $2 million in funding from KAUST Innovation Ventures and collaborated with KAUST to develop the technology. Why it matters: This innovation could significantly improve the battery life of mobile devices while also enhancing display quality, providing a competitive advantage for devices manufactured in the region.
KAUST · · Partnership Research
KAUST and LONGi, a solar energy company, have signed an MoU to collaborate on solar energy innovation. The partnership aims to develop and deploy solar technologies, focusing on improving the efficiency of photovoltaic technologies. Joint projects will include testing PV module concepts, developing new materials, and optimizing performance, alongside talent development programs. Why it matters: This collaboration aligns with Saudi Vision 2030 and the Saudi Green Initiative, fostering sustainable energy solutions and local expertise in the region.
This paper introduces a convolutional transformer model for classifying tomato maturity, along with a new UAE-sourced dataset, KUTomaData, for training segmentation and classification models. The model combines CNNs and transformers and was tested against two public datasets. Results showed state-of-the-art performance, outperforming existing methods by significant margins in mAP scores across all three datasets.
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.
KAUST researchers have discovered how sea anemones recycle nitrogen waste, allowing them to thrive in nutrient-poor ocean environments. The study used laser microdissection and single-cell RNA-sequencing to analyze tissue-specific gene expression in Aiptasia. They found that anemones distribute glucose received from symbionts across tissues to recycle nitrogen waste. Why it matters: This research enhances understanding of coral reef ecosystems and their resilience, which is particularly relevant for Red Sea biodiversity and Saudi Arabia's environmental efforts.
Researchers at KAUST, USTC, and SUSTech have developed a method for carbon capture and storage using guanidinium sulfate salt to create clathrate structures that trap CO2 molecules. This salt-based structure mimics methane hydrate activity and captures CO2 through physisorption, without water or nitrogen interference. The method allows CO2 to be carried as a solid powder at ambient temperature and pressure, offering a less energy-intensive alternative to traditional methods. Why it matters: This innovation introduces a new, energy-efficient way to store and transport CO2 as a solid, potentially revolutionizing carbon capture and storage technologies in the region and beyond.
KAUST · · Healthcare Research
KAUST and King Faisal Specialist Hospital and Research Centre (KFSH&RC) are collaborating to develop bioelectronic sensors for rapid pathogen detection. These sensors aim to provide cheap and accurate results, potentially replacing conventional lab tests. A COVID-19 saliva test developed by KAUST researchers showed comparable sensitivity to PCR tests with a 15-minute turnaround. Why it matters: This partnership accelerates the development of novel diagnostic tools, which could improve healthcare accessibility in remote areas and low-income countries within the region.
KAUST · · Research Infrastructure
KAUST researchers have developed a perovskite/silicon tandem solar cell with a power conversion efficiency (PCE) of 33.2%, surpassing the previous record of 32.5% held by Helmholtz Zentrum Berlin (HZB). The tandem device was certified by the European Solar Test Installation (ESTI) and listed at the top of the NREL efficiency chart. The cell combines perovskite top cells for blue light absorption with silicon bottom cells for red light absorption. Why it matters: This breakthrough could accelerate the adoption of high-performance photovoltaic modules, which is critical for achieving global renewable energy goals.
Arab News News · · Policy Infrastructure
Saudi Arabia and the UAE are actively positioning themselves as leading hubs for artificial intelligence development across the Middle East and North Africa region. Both nations are implementing comprehensive national AI strategies, investing significantly in advanced infrastructure, and fostering an environment conducive to technological innovation. These initiatives involve substantial government funding for AI research, talent acquisition, and strategic partnerships with global tech entities to build robust AI ecosystems. Why it matters: The concentrated efforts and financial commitments by Saudi Arabia and the UAE are pivotal in accelerating the region's digital transformation and establishing them as key players in the global AI landscape.
KAUST · · Infrastructure Research
Al-Miyah Solutions, a KAUST spin-off, has installed a decentralized wastewater treatment unit in Rabigh, Saudi Arabia. The mobile, modular technology converts wastewater into reusable water for areas lacking centralized sewer access, addressing a need for 40% of Saudi households. The technology, developed at KAUST's Water Desalination and Reuse Center, uses aerobic granular sludge and gravity-driven membrane filtration. Why it matters: This deployment offers a sustainable and cost-effective solution for wastewater treatment and reuse in remote areas of the Kingdom, reducing reliance on costly centralized systems.
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.
The paper introduces a novel actor-critic framework called Distillation Policy Optimization that combines on-policy and off-policy data for reinforcement learning. It incorporates variance reduction mechanisms like a unified advantage estimator (UAE) and a residual baseline. The empirical results demonstrate improved sample efficiency for on-policy algorithms, bridging the gap with off-policy methods.
KAUST · · Funding Partnership
KAUST startup Lihytech has raised US$6 million in funding from Ma'aden and the KAUST Innovation Ventures Fund. Lihytech's patented membrane technology, developed by Professor Zhiping Lai at KAUST, extracts battery-grade lithium from sources like seawater. The funding will be used to build a pilot facility at KAUST to extract lithium from the Red Sea and other in-Kingdom resources. Why it matters: This investment supports Saudi Arabia's goal of developing a complete electric vehicle value chain and becoming a key player in meeting global lithium demand.
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.
KAUST · · Partnership Research
KAUST, ENOWA, and SEC are partnering on a pilot project to capture 30 tonnes of CO2 per day from SEC’s Green Duba power plant at NEOM, using cryogenic carbon capture technology. The captured carbon will be used for food and beverage applications and to produce e-fuels. Saudi Electricity Company will build and operate the pilot plant. Why it matters: This partnership demonstrates Saudi Arabia's commitment to decarbonization and showcases the potential of cryogenic carbon capture for reducing the Kingdom's carbon footprint.
KAUST · · Research Partnership
Researchers from KAUST and Beneath The Waves used camera tags on tiger sharks in the Bahamas to map the world's largest seagrass ecosystem, estimated at 92,000 km2. The study, published in Nature Communications, details how the sharks' movements helped validate spatial estimates of these blue carbon ecosystems. Carlos Duarte (KAUST) realized the potential of the shark movement data to map the extensive seagrass meadows in the shallow sand banks of the Bahamas. Why it matters: The innovative approach demonstrates a novel method for mapping and quantifying vast underwater ecosystems, crucial for understanding and conserving blue carbon sinks.
KAUST · · Research Infrastructure
KAUST researchers have developed an ultrathin polymer-based membrane for water desalination with high water flux and salt rejection. The membrane utilizes two-dimensional porous carbonaceous materials with subnanometer-sized molecular transport channels. The membrane outperformed existing desalination systems using carbon nanotubes and graphene in forward and reverse osmosis. Why it matters: This innovation offers a promising alternative for efficient and cost-effective desalination, addressing critical water scarcity challenges in the region and beyond.
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.
KAUST · · Research Partnership
Matthew McCabe, director of the KAUST Climate and Livability Initiative (CLI), and his team have been awarded the 2022 Prince Sultan Bin Abdulaziz International Prize for Water in the Water Management and Protection category. The award recognizes their innovative use of satellites for water accounting and management, harmonizing data from CubeSat satellite platforms. They produced the highest resolution estimates of water usage ever retrieved from space, using data from Planet's constellation of small satellites. Why it matters: This award highlights the growing role of remote sensing technologies and KAUST's leadership in addressing critical climate and sustainability issues in water resource management within Saudi Arabia and globally.
KAUST · · Infrastructure Partnership
KAUST has selected HPE to build "Shaheen III", a next-generation supercomputer using the HPE Cray EX platform. Shaheen III will be 20 times faster than KAUST's existing system, making it the most powerful supercomputer in the Middle East. The system will support research in areas like clean combustion, Red Sea ecosystems, and climate events. Why it matters: This infrastructure investment will significantly boost AI and scientific computing capabilities in the region, enabling KAUST to tackle complex research challenges.
KAUST · · Research Partnership
KAUST, in partnership with NEOM, is launching the KAUST Reefscape Restoration Initiative at Shushah Island in the Red Sea. The project will restore approximately 100 hectares of reefscape by growing and planting hundreds of thousands of corals. It will also establish a research and ecotourism center. Why it matters: This initiative demonstrates Saudi Arabia's commitment to preserving coral reefs using advanced research and technology, which could have significant implications for marine ecosystem conservation in the region and globally.
arXiv · · CV Healthcare
Researchers from MBZUAI have developed EchoCoTr, a novel spatiotemporal deep learning method for estimating left ventricular ejection fraction (LVEF) from echocardiograms. EchoCoTr combines CNNs and vision transformers to overcome the limitations of each when applied to medical video data. The method achieves state-of-the-art results on the EchoNet-Dynamic dataset, demonstrating improved accuracy compared to existing approaches, with code available on GitHub.
KAUST · · Research Partnership
KAUST Associate Professor Peiying Hong has developed a wastewater treatment method using anaerobic membrane bioreactor (AnMBR) technology, which converts organic carbon into methane. In partnership with MODON, a pilot program is operational in Jeddah, treating 23,000 liters of wastewater daily using UV light and hydrogen peroxide for disinfection. This system produces clean water suitable for agriculture and biomass for fertilizer, with a smaller footprint and lower energy consumption than traditional aerobic methods. Why it matters: The AnMBR technology aligns with Saudi Vision 2030's water reuse objectives, reducing reliance on energy-intensive desalination and offering a sustainable solution for water-stressed regions.
KAUST · · Healthcare Research
KAUST researchers have developed a CRISPR-Cas system using a heat-stable Cas13 protein (TccCas13a) from Thermoclostridium caenicola, compatible with RT-LAMP for rapid viral detection. The new assay, named OPTIMA-dx, enhances the specificity of RT-LAMP tests by reducing false positives in SARS-CoV-2 detection. The team, led by Dr. Magdy Mahfouz and doctoral student Ahmed Mahas, is transitioning the product to a startup phase for commercialization. Why it matters: This innovation could significantly improve point-of-care diagnostics for COVID-19 and other infections by providing a more accurate and easier-to-use testing method.