This paper introduces Diffusion-BBO, a new online black-box optimization (BBO) framework that uses a conditional diffusion model as an inverse surrogate model. The framework employs an Uncertainty-aware Exploration (UaE) acquisition function to propose scores in the objective space for conditional sampling. The approach is shown theoretically to achieve a near-optimal solution and empirically outperforms existing online BBO baselines across 6 scientific discovery tasks.
KAUST Professor Aamir Farooq has been named a Fellow of the Royal Society of Chemistry (FRSC) for his contributions to chemical kinetics research. Farooq leads the KAUST Chemical Kinetics and Laser Sensors Laboratory, which focuses on understanding fundamental processes in energy conversion. His research currently investigates the chemistry of e-fuels, biofuels, low-carbon fuels, and zero-carbon fuels, with a focus on hydrogen and ammonia. Why it matters: This recognition highlights KAUST's contributions to sustainable energy research and positions the university as a key player in developing future fuel technologies.
Dr. Zeke Xie from HKUST(GZ) presented research on noise initialization and sampling strategies for diffusion models. The talk covered golden noise for text-to-image models, zigzag diffusion sampling, smooth initializations for video diffusion, and leveraging image diffusion for video synthesis. Xie leads the xLeaF Lab, focusing on optimization, inference, and generative AI, with previous experience at Baidu Research. Why it matters: The work addresses core challenges in improving the quality and diversity of generated content from diffusion models, a key area of advancement for AI applications in the region.
KAUST researchers are exploring novel chemical reactors and separation processes using mathematical design, with a focus on time and shape variables to enhance transport, heat transfer, and mass transfer. By aligning design, modeling, and 3D printing, they create customized shapes with great complexity and less material. This approach allows for the creation of bespoke reactors and separation processes tailored to specific applications, improving efficiency and reducing energy consumption. Why it matters: This research demonstrates the potential of advanced manufacturing techniques to revolutionize industrial design in the Middle East's chemical and pharmaceutical sectors.
KAUST researchers presented their work on stabilizing nanoparticle catalysts at the 252nd American Chemical Society Meeting & Exposition. The team devised a "molecular Scotch tape" using a silica gel support coated with a single molecule layer of soft material containing sulfur. This approach allows nanoparticles to stick to one side while leaving the other side free for catalysis, preventing aggregation without killing the catalyst. Why it matters: This innovation in catalyst stabilization could lead to more efficient and sustainable chemical processes, impacting various industries.
KAUST researchers have made several advances, including a new computational model of the Red Sea's ocean circulation. They also synthesized new metal-organic frameworks for gas storage with applications in green and medical tech. Additionally, they presented a mathematical solution for microgrid cybersecurity. Why it matters: These diverse research projects highlight KAUST's contributions to environmental modeling, materials science, and critical infrastructure protection in the region.
KAUST researchers developed a new single-molecule imaging method called the cumulative-area (CA) method. This method allows for simultaneous characterization of size, shape, and conformational dynamics of individual molecules, along with accurate determination of diffusion kinetics. The researchers demonstrated the CA method's effectiveness on nano- and micro-sized objects, extracting quantitative information about size, diffusion, and relaxation time. Why it matters: This advancement expands the capabilities of molecule imaging techniques in the region and has potential applications in polymer dynamics research and the study of molecular mechanisms within cells.
KAUST researchers studied quantum dot (QD) solar cells, finding that QD size significantly impacts electron injection efficiency. Using femtosecond broadband transient absorption spectroscopy, they examined charge transfer between QDs and phenyl-C61-butyric acid methyl ester (PCBM). They demonstrated that smaller QDs with a bandgap larger than 1 eV facilitate electron transfer to PCBM upon light absorption. Why it matters: This work provides insights into optimizing QD solar cell design by tuning electron injection through QD size, potentially leading to more efficient and low-cost photovoltaic technologies.