General AI research publications from MBZUAI, KAUST, TII, QCRI, and other GCC institutions, including conference papers, journal articles, and preprints.
Researchers at KAUST, Fraunhofer ISE, and University of Freiburg developed a method using 1,3-diaminopropane dihydroiodide (PDAI) to treat the perovskite surface of perovskite silicon tandem solar cells. The treated solar cells achieved a conversion efficiency of 33.1% and an open-circuit voltage of 2.01 volts. The devices maintained performance at over 40°C for over 1500 hours along the Saudi coast. Why it matters: This innovation overcomes challenges in surface passivation of textured perovskite cells, paving the way for more efficient and stable solar energy solutions suitable for deployment in hot climates.
The UAE has launched K2 Think, a new open-source artificial intelligence model. This model is being presented as the world's most advanced in reasoning capabilities. It is designed to offer sophisticated cognitive problem-solving to the global AI community. Why it matters: This launch underscores the UAE's strategic commitment to advancing cutting-edge AI research and development, providing a significant open-source tool for complex AI applications.
The paper introduces AraHalluEval, a new framework for evaluating hallucinations in Arabic and multilingual large language models (LLMs). The framework uses 12 fine-grained hallucination indicators across generative question answering and summarization tasks, evaluating 12 LLMs including Arabic-specific, multilingual, and reasoning-based models. Results show factual hallucinations are more common than faithfulness errors, with the Arabic model Allam showing lower hallucination rates. Why it matters: This work addresses a critical gap in Arabic NLP by providing a comprehensive tool for assessing and mitigating hallucination in LLMs, which is essential for reliable AI applications in the Arabic-speaking world.
Researchers from MBZUAI have introduced SPECS, a new reference-free evaluation metric for long image captions that modifies CLIP to emphasize specificity. SPECS aims to improve the correlation with human judgment while maintaining computational efficiency compared to LLM-based metrics. The proposed approach is intended for iterative use during image captioning model development, offering a practical alternative to existing methods.
The researchers introduce KAU-CSSL, the first continuous Saudi Sign Language (SSL) dataset focusing on complete sentences. They propose a transformer-based model using ResNet-18 for spatial feature extraction and a Transformer Encoder with Bidirectional LSTM for temporal dependencies. The model achieved 99.02% accuracy in signer-dependent mode and 77.71% in signer-independent mode, advancing communication tools for the SSL community.
This paper introduces AraDhati+, a new comprehensive dataset for Arabic subjectivity analysis created by combining existing datasets like ASTD, LABR, HARD, and SANAD. The researchers fine-tuned Arabic language models including XLM-RoBERTa, AraBERT, and ArabianGPT on AraDhati+ for subjectivity classification. An ensemble decision approach achieved 97.79% accuracy. Why it matters: The work addresses the under-resourced nature of Arabic NLP by providing a new dataset and demonstrating strong results in subjectivity classification, advancing sentiment analysis capabilities for the Arabic language.
This paper presents a UI-level evaluation of ALLaM-34B, an Arabic-centric LLM developed by SDAIA and deployed in the HUMAIN Chat service. The evaluation used a prompt pack spanning various Arabic dialects, code-switching, reasoning, and safety, with outputs scored by frontier LLM judges. Results indicate strong performance in generation, code-switching, MSA handling, reasoning, and improved dialect fidelity, positioning ALLaM-34B as a robust Arabic LLM suitable for real-world use.
The QU-NLP team presented their approach to the QIAS 2025 shared task on Islamic Inheritance Reasoning, fine-tuning the Fanar-1-9B model using LoRA and integrating it into a RAG pipeline. Their system achieved an accuracy of 0.858 on the final test, outperforming models like GPT 4.5, LLaMA, and Mistral in zero-shot settings. The system particularly excelled in advanced reasoning, achieving 97.6% accuracy. Why it matters: This demonstrates the effectiveness of domain-specific fine-tuning and retrieval augmentation for Arabic LLMs in complex reasoning tasks, even surpassing frontier models.
This paper introduces Saudi-Dialect-ALLaM, a LoRA fine-tuned version of the Saudi Arabian foundation model ALLaM-7B-Instruct-preview, designed to improve the generation of Saudi dialects (Najdi and Hijazi). The model is trained on a private dataset of 5,466 synthetic instruction-response pairs, with two variants explored: Dialect-Token and No-Token training. Results indicate that the Dialect-Token model achieves superior dialect control and fidelity compared to generic instruction models, although the dataset and model weights are not released.
Tajikistan has unveiled plans to establish Central Asia's first regional Artificial Intelligence (AI) center. The initiative aims to foster AI development, research, and collaboration across the Central Asian region. This center is expected to serve as a hub for innovation, knowledge exchange, and capacity building in AI technologies. Why it matters: This move could position Tajikistan as a leader in AI for Central Asia, significantly boosting regional technological advancement and potentially attracting international investment and expertise to the region.
This paper benchmarks the performance of large language models (LLMs) on Arabic medical natural language processing tasks using the AraHealthQA dataset. The study evaluated LLMs in multiple-choice question answering, fill-in-the-blank, and open-ended question answering scenarios. The results showed that a majority voting solution using Gemini Flash 2.5, Gemini Pro 2.5, and GPT o3 achieved 77% accuracy on MCQs, while other LLMs achieved a BERTScore of 86.44% on open-ended questions. Why it matters: The research highlights both the potential and limitations of current LLMs in Arabic clinical contexts, providing a baseline for future improvements in Arabic medical AI.
KAUST, Tufts, and JIHS researchers created pangenome graphs using Saudi and Japanese samples, named JaSaPaGe. These graphs address the underrepresentation of these populations in existing pangenome databases, which are used as references for understanding individual DNA. The population-specific pangenomes are expected to improve variant calling and diagnostic accuracy for genetic disorders in these groups. Why it matters: This work promotes precision medicine and reduces diagnostic gaps for underrepresented populations by providing more relevant genetic baselines.
A KAUST-led study in Nature proposes reversing land degradation by 2050 through increased sustainable seafood production, reduced food waste, and land restoration. The study suggests straightforward measures like modifying economic incentives and promoting sustainable aquaculture policies. Researchers estimate these policies could save a land area roughly the size of Africa. Why it matters: The KAUST-led research offers a tangible blueprint for addressing critical food security challenges in arid regions like Saudi Arabia and globally.
The UAE government is developing large language models (LLMs) specifically for the Arabic language, with a target training dataset of 20 million words. This initiative aims to overcome the underrepresentation of Arabic in existing AI models. The project seeks to enhance AI's ability to understand and generate nuanced Arabic content. Why it matters: A national Arabic LLM can enable culturally relevant AI applications across various sectors in the region, from education to government services.
KAUST researchers have developed an AI system for the Saudi Geological Survey (SGS) to improve the scientific understanding of seismic activity in Saudi Arabia. The AI system helps the SGS analyze swarm earthquakes, which are common in volcanic regions and difficult to decipher using conventional methods. The system allows for a more reliable survey of seismic regions, better infrastructure planning, and improved building codes. Why it matters: The AI system enhances Saudi Arabia's ability to monitor and respond to seismic events, contributing to public safety and infrastructure resilience.
Saudi Arabia's national AI authority, SDAIA (Saudi Data and AI Authority), is reportedly developing a National AI Index. This index aims to measure and track the Kingdom's progress and capabilities across various dimensions of artificial intelligence. It is expected to encompass areas such as research and development, talent, infrastructure, and ethical considerations. Why it matters: This initiative provides a crucial framework for evaluating Saudi Arabia's AI ecosystem, guiding future strategic investments, and benchmarking its national AI capabilities against international standards.
KAUST researchers found that sulfate ions reduce free water in aqueous batteries, mitigating parasitic reactions that degrade the anode and shorten battery life. Adding zinc sulfate increased battery lifespan by more than ten times. Sulfate salts stabilize the bonds of free water, acting as a "water glue" to reduce parasitic reactions. Why it matters: This finding provides a cheap and scalable approach to improve the viability of aqueous batteries for sustainable energy storage, particularly for integrating renewable energy sources.
Researchers introduce UnsafeChain, a new safety alignment dataset designed to improve the safety of large reasoning models (LRMs) by focusing on 'hard prompts' that elicit harmful outputs. The dataset identifies and corrects unsafe completions into safe responses, exposing models to unsafe behaviors and guiding their correction. Fine-tuning LRMs on UnsafeChain demonstrates enhanced safety and preservation of general reasoning ability compared to existing datasets like SafeChain and STAR-1.
This paper explores Dialectal Arabic (DA) to Modern Standard Arabic (MSA) machine translation using prompting and fine-tuning techniques for Levantine, Egyptian, and Gulf dialects. The study found that few-shot prompting outperformed zero-shot and chain-of-thought methods across six large language models, with GPT-4o achieving the highest performance. A quantized Gemma2-9B model achieved a chrF++ score of 49.88, outperforming zero-shot GPT-4o (44.58). Why it matters: The research provides a resource-efficient pipeline for DA-MSA translation, enabling more inclusive language technologies by addressing the challenges posed by dialectal variations in Arabic.
Researchers propose MS-NN-steer, a model-structured neural network for autonomous vehicle steering control that integrates nonlinear vehicle dynamics. The controller was validated using real-world data from the Abu Dhabi Autonomous Racing League (A2RL) competition. MS-NN-steer demonstrates improved accuracy, generalization, and robustness compared to general-purpose NNs and the A2RL winning team's controller. Why it matters: This research demonstrates a promising approach to developing transparent and reliable AI for safety-critical autonomous racing applications in the UAE.
KAUST is partnering with digiLab to develop AI for coral conservation within the KAUST Coral Restoration Initiative (KCRI). digiLab's AI platform will provide real-time simulations of the 100-hectare reefscape, aiding in understanding coral resilience and growth under changing conditions. The AI tools are expected to reduce coral assessment times from months to weeks and optimize sensor placement. Why it matters: This partnership sets a new standard for coral restoration by demonstrating a scalable AI-driven model for global conservation efforts.
KAUST researchers have developed deepBlastoid, a deep learning tool for evaluating models of human embryo development, called blastoids. deepBlastoid can evaluate images of blastoids at speeds 1000 times faster than expert scientists, processing 273 images per second. Trained on over 2000 microscopic blastoid images, it assesses the impact of chemicals on blastoid development using over 10,000 images. Why it matters: This AI tool accelerates research into early pregnancy, fertility complications, and the impact of chemicals on embryo development, with implications for reproductive technologies.
This paper introduces Absher, a new benchmark for evaluating LLMs' linguistic and cultural competence in Saudi dialects. The benchmark comprises over 18,000 multiple-choice questions spanning six categories, using dialectal words, phrases, and proverbs from various regions of Saudi Arabia. Evaluation of state-of-the-art LLMs reveals performance gaps, especially in cultural inference and contextual understanding, highlighting the need for dialect-aware training.
Researchers developed a two-stage AI pipeline to predict desalination performance efficiency losses due to climate factors in the UAE, achieving 98% accuracy. The model forecasts aerosol optical depth (AOD) and uses it to predict desalination efficiency, incorporating meteorological data. A dust-aware control logic was developed to optimize plant operations, and an interactive dashboard was created for decision support.
MBZUAI researchers have introduced MIRA, a novel framework for improving the factual accuracy of multimodal large language models in medical applications. MIRA uses calibrated retrieval to manage factual risk and integrates image embeddings with a medical knowledge base for efficient reasoning. Evaluated on medical VQA and report generation benchmarks, MIRA achieves state-of-the-art results, with code available on GitHub.
The paper introduces ScoreAdv, a novel approach for generating natural adversarial examples (UAEs) using diffusion models. It incorporates an adversarial guidance mechanism and saliency maps to shift the sampling distribution and inject visual information. Experiments on ImageNet and CelebA datasets demonstrate state-of-the-art attack success rates, image quality, and robustness against defenses.
The authors introduce Nile-Chat, a collection of LLMs (4B, 3x4B-A6B, and 12B) specifically for the Egyptian dialect, capable of understanding and generating text in both Arabic and Latin scripts. A novel language adaptation approach using the Branch-Train-MiX strategy is used to merge script-specialized experts into a single MoE model. Nile-Chat models outperform multilingual and Arabic LLMs like LLaMa, Jais, and ALLaM on newly introduced Egyptian benchmarks, with the 12B model achieving a 14.4% performance gain over Qwen2.5-14B-Instruct on Latin-script benchmarks; all resources are publicly available. Why it matters: This work addresses the overlooked aspect of adapting LLMs to dual-script languages, providing a methodology for creating more inclusive and representative language models in the Arabic-speaking world.
Researchers from MBZUAI introduce Forget-MI, a machine unlearning method tailored for multimodal medical data, enhancing privacy by removing specific patient data from AI models. Forget-MI utilizes loss functions and perturbation techniques to unlearn both unimodal and joint data representations. The method demonstrates superior performance in reducing Membership Inference Attacks and improving data removal compared to existing techniques, while preserving overall model performance and enabling data forgetting.
This paper introduces MOTOR, a multimodal retrieval and re-ranking approach for medical visual question answering (MedVQA) that uses grounded captions and optimal transport to capture relationships between queries and retrieved context, leveraging both textual and visual information. MOTOR identifies clinically relevant contexts to augment VLM input, achieving higher accuracy on MedVQA datasets. Empirical analysis shows MOTOR outperforms state-of-the-art methods by an average of 6.45%.
Researchers at KAUST and KACST have developed a composite material that enhances solar cell performance by absorbing air moisture at night and releasing it during the day. When applied to solar cells in Saudi Arabia, the material increased power output by 12.9% and extended cell lifespan by over 200%. The passive cooling technology also reduced electricity generation costs by 18%. Why it matters: This innovation addresses a key challenge in solar energy adoption in hot climates, potentially making solar power more efficient and cost-effective in the region.
A new methodology emulating fact-checker criteria assesses news outlet factuality and bias using LLMs. The approach uses prompts based on fact-checking criteria to elicit and aggregate LLM responses for predictions. Experiments demonstrate improvements over baselines, with error analysis on media popularity and region, and a released dataset/code at https://github.com/mbzuai-nlp/llm-media-profiling.
This paper benchmarks reasoning-focused LLMs, especially DeepSeek models, on fifteen Arabic NLP tasks. The study uses zero-shot, few-shot, and fine-tuning strategies. Key findings include that three in-context examples improve F1 scores by over 13 points on classification tasks, DeepSeek outperforms GPT-4-mini by 12 F1 points on complex inference tasks in the zero-shot setting, and LoRA fine-tuning yields up to an additional 8 points in F1 and BLEU. Why it matters: The systematic evaluation provides insights into the performance of LLMs on Arabic NLP, highlighting the effectiveness of different strategies for improving performance and contributing to the development of more capable Arabic language models.
A new benchmark, ViMUL-Bench, is introduced to evaluate video LLMs across 14 languages, including Arabic, with a focus on cultural inclusivity. The benchmark includes 8k manually verified samples across 15 categories and varying video durations. A multilingual video LLM, ViMUL, is also presented, along with a training set of 1.2 million samples, with both to be publicly released.
MBZUAI researchers introduce TerraFM, a scalable self-supervised learning model for Earth observation that uses Sentinel-1 and Sentinel-2 imagery. The model unifies radar and optical inputs through modality-specific patch embeddings and adaptive cross-attention fusion. TerraFM achieves strong generalization on classification and segmentation tasks, outperforming prior models on GEO-Bench and Copernicus-Bench.
Researchers from MBZUAI have introduced VideoMolmo, a large multimodal model for spatio-temporal pointing conditioned on textual descriptions. The model incorporates a temporal module with an attention mechanism and a temporal mask fusion pipeline using SAM2 for improved coherence across video sequences. They also curated a dataset of 72k video-caption pairs and introduced VPoS-Bench, a benchmark for evaluating generalization across real-world scenarios, with code and models publicly available.
MBZUAI researchers introduce VideoMathQA, a new benchmark for evaluating mathematical reasoning in videos, requiring models to interpret visual information, text, and spoken cues. The dataset spans 10 mathematical domains with videos ranging from 10 seconds to over 1 hour, and includes multi-step reasoning annotations. The benchmark aims to evaluate temporal cross-modal reasoning and highlights the limitations of existing approaches in complex video-based mathematical problem solving.
This paper introduces a novel evaluation framework for Arabic language models, addressing gaps in linguistic accuracy and cultural alignment. The authors analyze existing datasets and present the Arabic Depth Mini Dataset (ADMD), a curated collection of 490 questions across ten domains. Evaluating GPT-4, Claude 3.5 Sonnet, Gemini Flash 1.5, CommandR 100B, and Qwen-Max using ADMD reveals performance variations, with Claude 3.5 Sonnet achieving the highest accuracy at 30%. Why it matters: The work emphasizes the importance of cultural competence in Arabic language model evaluation, providing practical insights for improvement.
MBZUAI introduces Agent-X, a benchmark for evaluating multi-step reasoning in vision-centric agents across real-world, multimodal settings. Agent-X includes 828 tasks with diverse visual contexts and spans six environments, requiring tool use and stepwise decision-making. Experiments show that current LLMs struggle with multi-step vision tasks, achieving less than 50% success, highlighting areas for improvement in LMM reasoning and tool use.
This paper analyzes Arabic text generated by LLMs like ALLaM, Jais, Llama, and GPT-4 across academic and social media domains using stylometric analysis. The study found detectable linguistic patterns that differentiate human-written from machine-generated Arabic text. BERT-based detection models achieved up to 99.9% F1-score in formal contexts, though cross-domain generalization remains a challenge. Why it matters: The research lays groundwork for detecting AI-generated misinformation in Arabic, a crucial step for preserving information integrity in Arabic-language contexts.
MBZUAI researchers introduce SocialMaze, a new benchmark for evaluating social reasoning capabilities in large language models (LLMs). SocialMaze includes six diverse tasks across social reasoning games, daily-life interactions, and digital community platforms, emphasizing deep reasoning, dynamic interaction, and information uncertainty. Experiments show that LLMs vary in handling dynamic interactions, degrade under uncertainty, but can be improved via fine-tuning on curated reasoning examples.
KAUST researchers have developed a hybrid cooling technology combining nanotech plastic and biodegradable mulch that significantly enhances crop yields in arid regions. The technology lowers greenhouse temperatures by 25 degrees Celsius and doubles crop yields in tests with Chinese cabbage. The nanotech plastic coating absorbs infrared light, while the biodegradable mulch reflects sunlight to keep the soil cooler. Why it matters: This innovation promises to improve food security in arid regions like Saudi Arabia while reducing energy consumption and plastic waste associated with traditional greenhouse cooling methods.
Researchers from KAUST and KACST have developed a quantum random number generator (QRNG) that is almost 1000 times faster than existing QRNGs. The device utilizes micro-LEDs and advanced post-processing algorithms and has passed randomness tests by the National Institute of Standards and Technology. The QRNG's portability and high generation rate will benefit industries such as health, finance, and defense. Why it matters: This advancement significantly strengthens data security capabilities in Saudi Arabia, aligning with Vision 2030 goals for technological leadership and innovation.
The Qatar Computing Research Institute (QCRI) has released SpokenNativQA, a multilingual spoken question-answering dataset for evaluating LLMs in conversational settings. The dataset contains 33,000 naturally spoken questions and answers across multiple languages, including low-resource and dialect-rich languages. It aims to address the limitations of text-based QA datasets by incorporating speech variability, accents, and linguistic diversity. Why it matters: This benchmark enables more robust evaluation of LLMs in speech-based interactions, particularly for Arabic dialects and other low-resource languages.
MBZUAI researchers release 'Fann or Flop', a new benchmark for evaluating Arabic poetry understanding in LLMs. The benchmark covers 12 historical eras and 14 poetic genres, assessing semantic understanding, metaphor interpretation, and cultural context. Evaluation of state-of-the-art LLMs reveals challenges in poetic understanding despite strong performance on standard Arabic benchmarks.
MBZUAI researchers introduce ARB, the first comprehensive benchmark for evaluating step-by-step multimodal reasoning in Arabic across textual and visual modalities. The benchmark spans 11 diverse domains and includes 1,356 multimodal samples with 5,119 human-curated reasoning steps. Evaluations of 12 state-of-the-art LMMs revealed challenges in coherence, faithfulness, and cultural grounding, highlighting the need for culturally aware AI systems.
Researchers from MBZUAI have introduced UrduFactCheck, a new framework for fact-checking in Urdu, along with two datasets: UrduFactBench and UrduFactQA. The framework uses monolingual and translation-based evidence retrieval to address the lack of Urdu resources. Evaluations using twelve LLMs showed that translation-augmented methods improve performance, highlighting challenges for open-source LLMs in Urdu.
MBZUAI researchers introduce FAID, a fine-grained AI-generated text detection framework capable of classifying text as human-written, LLM-generated, or collaboratively written. FAID utilizes multi-level contrastive learning and multi-task auxiliary classification to capture authorship and model-specific characteristics, and can identify the underlying LLM family. The framework outperforms existing baselines, especially in generalizing to unseen domains and new LLMs, and includes a multilingual, multi-domain dataset called FAIDSet.
MBZUAI researchers introduce LLM-BabyBench, a benchmark suite for evaluating grounded planning and reasoning in LLMs. The suite, built on a textual adaptation of the BabyAI grid world, assesses LLMs on predicting action consequences, generating action sequences, and decomposing instructions. Datasets, evaluation harness, and metrics are publicly available to facilitate reproducible assessment.
Saudi health innovators are increasing investment in smart health solutions, using KAUST's infrastructure and expertise. The King Salman Center for Disability Research (KSCDR) and KCSH are partnering on AI-based methods to identify genetic causes of rare eye diseases. KAUST and the Saudi Food and Drug Administration (SFDA) are expanding their agreement to enhance cooperation on AI and digital innovation. Why it matters: These partnerships signal a concerted effort to leverage AI for addressing critical healthcare challenges and advancing the Kingdom's health priorities.
This paper introduces DaringFed, a novel dynamic Bayesian persuasion pricing mechanism for online federated learning (OFL) that addresses the challenge of two-sided incomplete information (TII) regarding resources. It formulates the interaction between the server and clients as a dynamic signaling and pricing allocation problem within a Bayesian persuasion game, demonstrating the existence of a unique Bayesian persuasion Nash equilibrium. Evaluations on real and synthetic datasets demonstrate that DaringFed optimizes accuracy and convergence speed and improves the server's utility.