A fluorinated imine additive improves graphite anode stability through a LiF-rich protective layer, enhancing lithium-ion transport and long-term battery stability. The additive retains 89.4% and 95.6% of maximum capacities after 1,000 cycles, outperforming additive-free cells.
A new probabilistic method, Bayesian Probabilistic Data Association via Gaussian Mixture Models, improves trajectory accuracy and semantic mapping quality in robots. The approach reduces duplicate registrations and handles ambiguous observations, enabling more stable and reliable object-level maps.
Researchers developed a novel divide-and-conquer approach for model checking linear temporal properties, called DCA2MC, to address state-space explosion and long verification times. The approach divides the original model checking problem into smaller, independent tasks, reducing memory consumption and verification time.
A study confirms the 20-item prosopagnosia index questionnaire is an effective screening tool for detecting developmental prosopagnosia, a condition where individuals struggle to recognize familiar faces. The researchers also identified opportunities for refinement to improve the questionnaire's accuracy.
UniSpec delivers lossless LLM acceleration without retraining while adapting automatically to different hardware platforms and multilingual workloads. The framework achieves up to 2.6× faster inference than existing methods across multiple models, hardware, and languages.
A new AI model developed by Japan Advanced Institute of Science and Technology identifies the most relevant information in videos, significantly reducing analysis time. The model achieved state-of-the-art performance while using only 15.42% of available visual features and reduced processing time per video by approximately 65%.
Researchers at Japan Advanced Institute of Science and Technology developed a new computational method combining neural networks with Bayesian localization, achieving accurate predictions while reducing computational cost. This breakthrough enables the high-precision analysis of large-scale materials and complex chemical reaction systems.
Researchers at JAIST and NIMS have developed a high-throughput screening strategy that simultaneously explores catalysts and reactions. The approach resulted in the discovery of promising materials and revealed minor products, including 1-butene and benzene, indicating early signs of unknown reaction pathways.
Researchers developed EleTac, a soft robotic gripper with high-resolution tactile sensing, to handle delicate objects. The gripper's innovative design enables it to adapt to various shapes and provide gentle forces, making it suitable for applications such as handling fruit, lab samples, and medical supplies.
Researchers developed ADASPEC to speed up multilingual AI systems by dynamically adapting to different languages during inference. The framework generates instruction data in any desired language using the target LLM itself, reducing unnecessary vocabulary computations and achieving faster and more stable multilingual inference.
Researchers design polymer networks to replicate dynamic behaviors inspired by biological systems. Self-oscillating gels exhibit rhythmic motion similar to a beating heart, while artificial photosynthetic gels convert light into chemical energy.
Researchers developed a method to separate and quantify proton transport at individual interfaces in ultrathin ionomer films, enabling the evaluation of interfacial transport properties. The study revealed that proton transport at different interfaces is of a similar order of magnitude.
Researchers develop a smarter image-generation system that produces accurate, high-quality building designs. The framework uses real architectural datasets to generate detailed sketches and renderings, overcoming previous limitations.
A new generative AI framework predicts future urban layouts considering multiple factors. The Memory-aware Multi-Conditional generation Network (MMCN) offers a novel approach to modeling complex urban evolution, providing a powerful tool for sustainable urban development.
Researchers developed Neuronal Type Assignment from Connectivity (NTAC) to accurately assign neuronal cell types based on synaptic wiring patterns. NTAC outperformed traditional morphology-based approaches in identifying neuron types, especially in complex brain regions.
Researchers develop AI framework to accelerate alloy discovery by fusing cross-disciplinary expert knowledge with experimental data, outperforming conventional machine learning methods. The approach can make reliable predictions for poorly studied alloy compositions, achieving accuracy rates up to 92%.
Researchers propose parametric urban design as a solution to improve urban design practice. This approach generates multiple possible layouts and assesses their impact on cardiometabolic health through algorithmic models. It addresses limitations of current methods by considering interdependencies between built environment features.
Researchers discovered that a bacterium isolated from Japanese tree frogs has complete tumor-eliminating properties. The study found that this bacterium attacks cancer through two mechanisms, selectively accumulating in tumors and evading the immune system.
Researchers discovered P. angustum selectively targets colorectal cancer, inducing direct tumor lysis and robust immune activation. The therapy promotes intratumoral infiltration of immune cells and enhances production of inflammatory cytokines, significantly prolonging survival in treated mice.
Advanced electron microscopy technique uncovers phase shifts in lithium battery cathodes, revealing spinel- and rocksalt-type structures that contribute to degradation. The study guides the design of longer-lasting batteries with higher energy densities.
A team of Japanese researchers has uncovered the deformation processes that give Kanazawa gold leaf its remarkable thinness and brilliance. The study used electron microscopy to reveal the activation of a rare crystal slip system, providing insights into the traditional crafting technique.
The research team developed multifunctional nanocomposites that demonstrate excellent tumor-targeting capability through the EPR effect. Irradiation with near-infrared laser light achieved multidimensional therapeutic effects, including complete elimination of transplanted mouse cancers within 5 days.
Researchers developed ProTac, a novel vision-based soft sensing skin for robots, enabling real-time environmental perception and dual-mode sensing. The system can detect approaching objects from multiple angles and recognize multiple touch points with high accuracy.
Researchers developed a new speculative decoding framework called SPECTRA, which optimizes internal and external speculation for accelerated text generation. By integrating a core module and an optional retrieval module, SPECTRA achieves substantial speedups across diverse tasks and model architectures while preserving output quality.
Maude-NPA's parallelization significantly reduces analysis time for complex cryptographic protocols. The new method improves runtime performance by an average of 52% and enables formal analysis of quantum-resistant TLS protocols.
Researchers have discovered a novel microbial consortium called AUN that produces exceptional tumor eradication in both murine and human cancer models. The therapy exhibits high biocompatibility and minimal side effects, offering a long-awaited solution for immunocompromised patients.
A recent study at JAIST uncovers the mechanisms behind symmetry breaking during meniscus splitting in evaporating polymer solutions. The experiment reveals that nucleation points form at uneven positions along the confined space, influencing the timing and positioning of subsequent splits.
Researchers developed a novel approach to automated spoken English assessment integrating acoustic, turn-taking, linguistic, and visual components. The multioutput learning framework achieved an accuracy of approximately 83% in predicting SEE scores.
Researchers used a new verification framework to test the safety of Autoware, revealing potential limitations in critical traffic situations. The study found that Autoware failed to consistently follow safety rules during scenarios like cut-in, cut-out, and deceleration, highlighting the need for improvement before real-world deployments.
Researchers introduce Onset Intensity for Temperature Dependence (OITD) to identify rate-limiting steps in photocatalytic reactions. The study reveals distinct rate-limiting behaviors for different materials, highlighting the importance of surface accessibility in optimizing photocatalytic material design.
Proof scores use term rewriting to verify system properties, striking a balance between automation and human effort. Despite limitations, the technique has been successfully applied to various systems and protocols, with potential for critical applications in safety-critical systems.
Researchers at JAIST have developed a low-dose imaging technique that maps the three-dimensional atomic structure of titanium oxyhydroxide nanoparticles without damaging them. This breakthrough enables safer analysis and opens possibilities for designing materials with enhanced functionality.
Researchers developed bacteria-enhanced graphene oxide nanoparticles that effectively destroy tumors through a three-pronged mechanism. The nanocomposites combine chemotherapy, immune activation, and photothermal heating to suppress tumor growth and activate strong immune responses in mice.
Researchers visualized the dynamic shuttling of α-CD rings along a PEG chain in real time, revealing localized structural changes. The study introduces a new method for analyzing supramolecular polymers and could pave the way for energy-efficient molecular motors.
A team of researchers from Japan Advanced Institute of Science and Technology proved that Dudeney's original solution to the famous dissection problem is the optimal solution. They used a novel approach, including matching diagrams, to show that no dissection between an equilateral triangle and a square can be achieved with three or fe...
Magnetic nanoparticles are guided to tumors using a magnet and heated by a laser to destroy cancer cells. Researchers developed nanoparticles that outperform conventional photothermal agents, killing cancer cells with high efficiency.
A recent study reveals three distinct mechanisms of recombination in photocatalytic water splitting, including over-penetration induced recombination and excess hole induced recombination. The discovery of a previously unknown slow reaction, called the 'satellite peak,' is crucial for pinpointing the rate-limiting step in water splitting.
Researchers at Japan Advanced Institute of Science and Technology developed Leafbot, a soft robot that uses vibration-driven locomotion to traverse uneven surfaces. The robot's compliant structure and simple motion strategy enable it to overcome complex obstacles, making it valuable for applications such as inspection and exploration.
Researchers developed Concurrent Dynamic Quantum Logic (CDQL) to verify quantum protocols with concurrent actions, enhancing expressiveness and speeding up verification. CDQL provides a rigorous framework for verifying both sequential and concurrent models of quantum protocols.
A novel framework for retrieving concise entailing legal articles has been proposed, achieving state-of-the-art results across two datasets. The Retrieve–Revise–Refine framework combines small and large language models to improve precision while limiting recall loss.
Researchers design bioinspired hydrogels that mimic plant photosynthesis for clean hydrogen energy production. The study achieves significant boosts in the activity of water-splitting processes and produces more hydrogen compared to older techniques.
A team of researchers created PenGym, an effective and reliable realistic training framework for RL pentesting agents that enables actual actions on realistic hosts in network environments. This approach yields promising results compared to simulated environments.
Researchers at Japan Advanced Institute of Science and Technology have developed a novel method to culture antitumor bacteria using porous scaffolds, enhancing their anticancer properties and improving safety in animal testing. The approach resulted in improved survival rates in mice with breast cancer, including drug-resistant cases.
Researchers developed ROSE, a soft robotic gripper that gently grasps objects using a unique 'wrinkling' phenomenon. The study demonstrates ROSE's effectiveness in picking up various crops, including strawberries and mushrooms, with high success rates.
Researchers at Japan Advanced Institute of Science and Technology developed a densely functionalized polymeric binder for high-performance lithium and sodium-ion batteries. The new material showed exceptional electrochemical performance, high capacities, and great cycle stability.
Researchers developed tumor cell-coated carbon nanohorns to deliver paclitaxel to colon cancer, exhibiting high accumulation at tumors and strong chemotherapeutic effects. The treatment also demonstrated a robust photothermal effect and immune responses, effectively destroying tumors.
Researchers have created a novel system called ConTac, which can estimate the shape and contact of a robotic arm with soft skin using a single sensing module. The system consists of a backbone, soft skin with markers, a camera to observe skin deformation, and models for shape and contact sensing.
A novel peptide, KS-133, has been developed to target genetic mechanisms associated with schizophrenia. The nanoparticle-based drug delivery system, where KS-133 is encapsulated with a brain-targeting peptide, shows effective distribution in the brains of mice and improves cognitive functions in mice with induced schizophrenia.
Researchers developed a symbolic model checking approach to verify quantum circuits, addressing the gap between model-checking quantum programs and quantum circuits. They used Maude programming language to formally specify and verify quantum circuits, confirming their correctness and paving the way for error-free quantum computing.
Researchers identify gaps in understanding how workplace design influences employee active and sedentary behaviors. A new study suggests that accurately measuring behaviors and analyzing spatial layouts can inform the design of work environments that facilitate engagement with inactive behaviors.