General AI research publications from MBZUAI, KAUST, TII, QCRI, and other GCC institutions, including conference papers, journal articles, and preprints.
KAUST and the Saudi Ministry of Environment, Water and Agriculture (MEWA) are collaborating on the Aquaculture Development Program (ADP) to advance Saudi Arabia's food security goals under Vision 2030. The ADP aims to increase domestic seafood production to 530,000 tons annually by 2030 through sustainable aquaculture practices. KAUST is employing a multidisciplinary team and innovative approaches like Integrated Multitrophic Aquaculture (IMTA) to optimize resource use and minimize environmental impact. Why it matters: This partnership aims to transform Saudi Arabia's aquaculture sector, reducing reliance on imports and promoting economic diversification while preserving marine biodiversity.
This paper introduces a deep vision-based framework for predicting coastal floods under climate change, addressing the challenges of limited training data and high-dimensional output. The framework employs and compares various deep learning models, including a custom compact CNN architecture, against geostatistical and traditional machine learning methods. A new synthetic dataset of flood inundation maps for Abu Dhabi's coast is also provided to benchmark future models.
KAUST spinout iyris, an AgriClimate Tech company, raised $16 million in Series A funding led by Ecosystem Integrity Fund. The funding will help iyris scale the sales and delivery of its SecondSky greenhouse covers and nets internationally. iyris' SecondSky technology was developed at KAUST and increases crop yields while reducing input costs. Why it matters: This funding highlights the potential of KAUST-backed startups to address critical challenges in agriculture and sustainability, particularly in harsh environments.
KAUST researchers discovered a five-hectare bio-sedimentary formation of living stromatolites off Sheybarah Island in the Red Sea. These structures are microbial carbonates similar to fossils of early life and are only the second group found in normal marine settings. The stromatolites host a diverse microbial community, including reticulated filaments previously only found in caves. Why it matters: The discovery provides insights into early life on Earth and has implications for understanding potential life formation on Mars, while also creating a unique educational opportunity for tourism in Saudi Arabia.
The UAE and France have officially signed a Memorandum of Understanding (MoU) to enhance cooperation in the field of artificial intelligence. This agreement establishes a framework for facilitating joint initiatives, knowledge exchange, and collaborative projects between their respective AI ecosystems. The specific areas of cooperation are expected to span various aspects of AI development and application across both nations. Why it matters: This MoU signifies a strategic bilateral partnership, potentially accelerating innovation, research, and policy alignment in AI between the UAE and France.
This survey paper reviews recent literature on continual learning in medical imaging, addressing challenges like catastrophic forgetting and distribution shifts. It covers classification, segmentation, detection, and other tasks, while providing a taxonomy of studies and identifying challenges. The authors also maintain a GitHub repository to keep the survey up-to-date with the latest research.
This paper introduces SimulMask, a new paradigm for fine-tuning large language models (LLMs) for simultaneous translation. SimulMask utilizes a novel attention masking approach that models simultaneous translation during fine-tuning by masking attention for a desired decision policy. Applied to a Falcon LLM on the IWSLT 2017 dataset, SimulMask achieves improved translation quality compared to state-of-the-art prompting optimization strategies across five language pairs while reducing computational cost. Why it matters: The proposed method offers a more efficient way to adapt LLMs for real-time translation, potentially enhancing multilingual communication tools and services.
Researchers from MBZUAI have introduced the Complex Video Reasoning and Robustness Evaluation Suite (CVRR-ES) for assessing Video-LLMs. The benchmark evaluates models across 11 real-world video dimensions, revealing challenges in robustness and reasoning, particularly for open-source models. A training-free Dual-Step Contextual Prompting (DSCP) technique is proposed to enhance Video-LMM performance, with the dataset and code made publicly available.
KAUST and Saudi Aramco collaborated to develop a laser-based sensor for detecting trace amounts of gas leaks in petrochemical plants. The sensor uses machine learning to identify specific gases, differentiating it from previous sensors that only detect large leaks. The technology can differentiate between closely related industrial gases like benzene, toluene, ethyl benzene and xylene (BTEX). Why it matters: This innovation enables proactive monitoring and rapid pinpointing of leaks, enhancing safety, environmental protection, and operational efficiency in the petrochemical industry.
Researchers compiled a 101 Billion Arabic Words Dataset by mining text from Common Crawl WET files and rigorously cleaning and deduplicating the extracted content. The dataset aims to address the scarcity of original, high-quality Arabic linguistic data, which often leads to bias in Arabic LLMs that rely on translated English data. This is the largest Arabic dataset available to date. Why it matters: The new dataset can significantly contribute to the development of authentic Arabic LLMs that are more linguistically and culturally accurate.
KAUST and King Faisal Specialist Hospital and Research Centre (KFSHRC) are collaborating to develop an RNA sequencing tool to improve the diagnosis rate of genetic diseases. The tool analyzes RNA data to find aberrant transcripts and mutations, building on KFSHRC's clinical data and KAUST's computational expertise. The team has already solved cases that DNA sequencing alone could not, including a case of a young child with brain damage caused by a recessive gene mutation. Why it matters: This collaboration can improve disease management and preventative services in the region, directly contributing to Saudi Arabia’s national research priority of health and wellness.
A study compared the vulnerability of C programs generated by nine state-of-the-art Large Language Models (LLMs) using a zero-shot prompt. The researchers introduced FormAI-v2, a dataset of 331,000 C programs generated by these LLMs, and found that at least 62.07% of the generated programs contained vulnerabilities, detected via formal verification. The research highlights the need for risk assessment and validation when deploying LLM-generated code in production environments.
KAUST has launched the KAUST Coral Restoration Initiative (KCRI), the world's largest coral restoration project, with a nursery on the NEOM coast capable of producing 40,000 corals annually. A secondary facility is under construction, designed to nurture 400,000 corals annually and expected to be completed by December 2025. The initiative aligns with Saudi Vision 2030 to bolster marine conservation efforts. Why it matters: This project demonstrates the Kingdom's commitment to environmental sustainability and leverages KAUST's research capabilities to address the critical issue of coral reef degradation, which has far-reaching implications for marine biodiversity and coastal communities.
KAUST and King Abdulaziz University (KAU) are collaborating to develop low-cost sodium-ion battery technology using fly ash, a waste material from burning fossil fuels. Researchers are purifying fly ash and using thermal treatment to engineer its structure for use as carbon electrodes in batteries. The resulting carbon electrode material is competitive with existing market products and can be used for other applications. Why it matters: This research offers a sustainable approach to energy storage by repurposing waste materials, potentially enabling cheaper and more environmentally friendly grid-scale energy storage for renewable energy sources.
Researchers at MBZUAI have developed DynaMMo, a dynamic model merging method for efficient class incremental learning using medical images. DynaMMo merges multiple networks at different training stages using lightweight learnable modules, reducing computational overhead. Evaluated on three datasets, DynaMMo achieved a 10-fold reduction in GFLOPS compared to existing dynamic methods with a 2.76 average accuracy drop.
KAUST researchers have developed an energy-efficient wastewater treatment process that generates high-quality effluent suitable for reuse. A pilot plant in Jeddah, operating since July 2022 in collaboration with MODON, treats 50,000 liters of wastewater daily off-grid, generating 1.5 kWh of electrical energy per 1,000 liters treated. The plant utilizes an anaerobic membrane bioreactor (AnMBR) coupled with UV disinfection, removing up to 99.9999% of microorganisms and producing less solid waste. Why it matters: This decentralized, energy-independent system offers a sustainable solution for water treatment in resource-scarce regions of the Middle East, aligning with Saudi Arabia's sustainability goals.
Researchers introduce ArabicaQA, a large-scale dataset for Arabic question answering, comprising 89,095 answerable and 3,701 unanswerable questions. They also present AraDPR, a dense passage retrieval model trained on the Arabic Wikipedia. The paper includes benchmarking of large language models (LLMs) for Arabic question answering. Why it matters: This work addresses a significant gap in Arabic NLP resources and provides valuable tools and benchmarks for advancing research in the field.
The paper introduces MedPromptX, a clinical decision support system using multimodal large language models (MLLMs), few-shot prompting (FP), and visual grounding (VG) for chest X-ray diagnosis, integrating imagery with EHR data. MedPromptX refines few-shot data dynamically for real-time adjustment to new patient scenarios and narrows the search area in X-ray images. The study introduces MedPromptX-VQA, a new visual question answering dataset, and demonstrates state-of-the-art performance with an 11% improvement in F1-score compared to baselines.
Researchers at MBZUAI introduce FissionFusion, a hierarchical model merging approach to improve medical image analysis performance. The method uses local and global aggregation of models based on hyperparameter configurations, along with a cyclical learning rate scheduler for efficient model generation. Experiments show FissionFusion outperforms standard model souping by approximately 6% on HAM10000 and CheXpert datasets and improves OOD performance.
Researchers at MBZUAI have introduced TiBiX, a novel approach leveraging temporal information from previous chest X-rays (CXRs) and reports for bidirectional generation of current CXRs and reports. TiBiX addresses two key challenges: generating current images from previous images and reports, and generating current reports from both previous and current images. The study also introduces a curated temporal benchmark dataset derived from the MIMIC-CXR dataset and achieves state-of-the-art results in report generation.
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.
KAUST researchers have developed a green synthetic biology approach using engineered algae to replicate the complex fragrances of agarwood, also known as oudh. They catalogued the chemical diversity of sesquiterpenes (STPs) in 58 agarwood samples and reproduced some of the chemical complexity of agarwood STPs in algae using synthetic biology. The team used the green alga Chlamydomonas reinhardtii to produce nine distinct STP chemical products widely found in agarwood, offering a sustainable alternative to harvesting endangered trees. Why it matters: This research provides a sustainable route for producing sought-after fragrances, reducing pressure on endangered agarwood tree populations and promoting green chemistry in the region.
Researchers at MBZUAI have introduced MedMerge, a transfer learning technique that merges weights from independently initialized models to improve performance on medical imaging tasks. MedMerge learns kernel-level weights to combine features from different models into a single model. Experiments across various medical imaging tasks demonstrated performance gains of up to 7% in F1 score.
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.
This paper introduces a new Single Domain Generalization (SDG) method called ConDiSR for medical image classification, using channel-wise contrastive disentanglement and reconstruction-based style regularization. The method is evaluated on multicenter histopathology image classification, achieving a 1% improvement in average accuracy compared to state-of-the-art SDG baselines. Code is available at https://github.com/BioMedIA-MBZUAI/ConDiSR.
Researchers from MBZUAI have developed XReal, a diffusion model for generating realistic chest X-ray images with precise control over anatomy and pathology location. The model utilizes an Anatomy Controller and a Pathology Controller to introduce spatial control in a pre-trained Text-to-Image Diffusion Model without fine-tuning. XReal outperforms existing X-ray diffusion models in realism, as evaluated by quantitative metrics and radiologists' ratings, and the code/weights are available.
KAUST researchers have developed a new synthetic biology process using metabolically engineered algae to produce fragrant sesquiterpenoids, the core compounds in agarwood and other perfumes. The process, developed by the Lauersen and Szekely groups, achieved yields 25 times higher than previous methods and allows for the synthesis of 103 types of fragrant sesquiterpenoids. It also incorporates an energy-efficient nanofiltration step and operates at room temperature with minimal waste. Why it matters: This sustainable bioprocess offers a green alternative to environmentally damaging harvesting of natural resources for the $44 billion fragrance industry, with potential applications in drug development.
KAUST scientists developed a new perovskite solar cell design using thin perovskite layers at the top and bottom of the interface. The new design achieves a power conversion efficiency of 25.6%, comparable to silicon solar cells, with only a 5% efficiency loss after 1000 hours of high heat exposure. The key innovation is the use of a specific ligand that interacts effectively with the 3D perovskites for passivation, maintaining purity in the thin layers. Why it matters: This advancement enhances the stability and efficiency of perovskite solar cells, making them a more viable and cost-effective alternative to silicon, especially for countries like Saudi Arabia aiming to increase renewable energy reliance.
KAUST researchers have developed new techniques to produce nutritious microalgae in industrial volumes using seawater-adapted Spirulina and Chlorella strains. This innovation eliminates the need for freshwater, making algae-based livestock feed production sustainable and economically viable. The new Saudi Center for Algal Biotechnology Development and Aquaculture will scale up operations from a 1,000 square meter pilot plant to 42,000 square meters. Why it matters: This could help Saudi Arabia decrease its dependency on imported feed and raw food materials, aligning with its Vision 2030 goals for increased domestic food security.
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.
KAUST researchers published a roadmap in Science for commercializing perovskite/silicon tandem solar cells, combining efficient light absorption with long-term stability. The roadmap addresses key challenges like real-world condition variability via geographical testing, accelerated stability testing, and high material costs. The researchers propose paradigms for both monolithic and mechanically stacked tandem production lines to identify cost-reduction opportunities. Why it matters: This work paves the way for affordable, accessible clean energy in Saudi Arabia and the world, with the market for perovskite/silicon tandems expected to exceed $10 billion within a decade.
KAUST researchers, in collaboration with Spanish scientists, have released the Global Ocean Gene Catalog 1.0, the world's largest open-source catalog of marine microbes. The catalog, created using the KAUST Metagenomic Analysis Platform (KMAP), matches microbial class with gene function, geographic location, and habitat type, including 317 million unique gene clusters. The catalog analyzes 2102 ocean samples taken from different depths and locations around the world. Why it matters: This resource will enable researchers to investigate ocean ecosystems, track pollution impact, and explore biotechnology applications, potentially driving significant advances in fields like antibiotic discovery and plastic degradation.
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.
Aramco and KAUST have launched the ENERCOMP consortium, a five-year initiative focused on R&D in nonmetallics and composites for energy applications. Aramco is the founding member and first research sponsor. The consortium aims to develop less energy-intensive and lower carbon-footprint materials, aligning with Saudi Arabia's sustainability vision. Why it matters: The partnership signals a strategic push towards diversifying the Kingdom's economy and strengthening its position in the energy and materials transitions, leveraging AI and advanced materials research.
KAUST researchers have made the first direct observations of active hydrothermal vent fields in the Red Sea, specifically on Hatiba Mons. The total active area spans 1.6 km2, the largest reported globally, with vent temperatures around 40 degrees Celsius supporting thriving microbial communities. The observations, confirmed using remotely operated underwater vehicles (ROVs), build on over a decade of data collection. Why it matters: These findings provide insights into the Red Sea's unique biological and mineralogical resources, offering a model to study the evolution of life in extreme environments.
Researchers introduce TomFormer, a transformer-based model for accurate and early detection of tomato leaf diseases, with the goal of deployment on the Hello Stretch robot for real-time diagnosis. TomFormer combines a visual transformer and CNN, achieving state-of-the-art results on KUTomaDATA, PlantDoc, and PlantVillage datasets. KUTomaDATA was collected from a greenhouse in Abu Dhabi, UAE.
The study introduces the Qatar University Dual-Machine Bearing Fault Benchmark dataset (QU-DMBF) containing sound and vibration data from two motors across 1080 conditions. It proposes a deep learning approach for sound-based fault detection, addressing limitations of vibration-based methods. Experiments on QU-DMBF show sound-based detection is more robust, independent of sensor location, and cost-effective while matching vibration-based performance. Why it matters: The new dataset and findings could shift the focus toward sound-based methods for more reliable and accessible predictive maintenance in industrial settings.
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.
ClimateCrete, a KAUST spinout, has raised investment for its technology that modifies sand particles to make them suitable for concrete manufacturing. The patented tech reduces the need for cement and lowers CO2 emissions by up to 60 percent. Tests show a significant increase in strength compared to untreated sand. Why it matters: This technology addresses the global shortage of suitable construction sand and supports Saudi Arabia's carbon-neutrality goals by enabling a 100% domestic supply chain for concrete.
KAUST researchers have found that dust clouds in the Arabian Peninsula are three times larger than previously estimated by current models. The study, published in the Journal of Geophysical Research: Atmospheres, uses refined mathematical models and data collected since 2012 to analyze the impact of coarse dust particles. The updated model indicates that larger particles contribute to over 80% of dust mass on land, leading to significant efficiency loss for solar technology, estimated at 15-45% depending on location. Why it matters: Accurate dust modeling is crucial for the strategic deployment and maintenance of solar technology, supporting Saudi Arabia's sustainable economy goals.
Researchers propose a universal anatomical embedding (UAE) framework for medical image analysis to learn appearance, semantic, and cross-modality anatomical embeddings. UAE incorporates semantic embedding learning with prototypical contrastive loss, a fixed-point-based matching strategy, and an iterative approach for cross-modality embedding learning. The framework was evaluated on landmark detection, lesion tracking and CT-MRI registration tasks, outperforming existing state-of-the-art methods.
Researchers at MBZUAI have developed GeoChat, a new vision-language model (VLM) specifically designed for remote sensing imagery. GeoChat addresses the limitations of general-domain VLMs in accurately interpreting high-resolution remote sensing data, offering both image-level and region-specific dialogue capabilities. The model is trained on a novel remote sensing multimodal instruction-following dataset and demonstrates strong zero-shot performance across tasks like image captioning and visual question answering.