Invention Description Retrieval-Augmented Generative (RAG) models boost generative AI’s accuracy by connecting large language models (LLMs) with the most current, external, knowledge sources. RAG models are widely used in fact-checking, information retrieval, and AI-driven search engines. Despite their utility, adversarial threats can exploit...

Invention Description Conducting learning-based models on large-scale unstructured geometric data, particularly volumetric 3D meshes, remains a significant challenge. While previous methods have employed graph convolutional approaches, they suffer from high space complexity, limiting efficient mini-batch training, as well as fixed short-range feature...

Invention Description Protein allostery is important in many biological processes and drug development. Understanding the mechanisms involved would be highly beneficial for therapeutic interventions. Existing techniques focus primarily on equilibrium dynamics and network analysis for allosteric interaction mapping. Unfortunately, these are limited...

Invention Description Viral evolution, particularly rapid viral evolution, can lead to the emergence of variants of concern. In order to develop effective antiviral strategies, it is essential to understand the molecular mechanisms driving these adaptations. Some research has focused on identifying adaptive mutations and how they impact protein structure...

Invention Description Type 1 diabetes (T1D) is an autoimmune disease where the body attacks and destroys insulin-producing beta cells in the pancreas and affects nearly 10 million people globally. Because glucose needs to be continuously monitored for precise insulin dosing, managing T1D is challenging. While automated insulin delivery (AID) systems...

Continuous-time reinforcement learning (CT-RL) algorithms hold great promise in real-world control applications. Adaptive dynamic programming (ADP)-based CT-RL algorithms, especially their theoretical developments, have achieved great successes. However, these methods have not been demonstrated for solving realistic or meaningful learning control problems....

Background Ground-based snow monitoring stations have long supported water management, forecasting, and planning. Recent advancements in remote sensing technology have improved the ability to monitor snow cover at broader scales. Satellite imagery, such as that from the Moderate Resolutiion Imaging Spectroradiometer (MODIS), have been used to track...

Event-based cameras that asynchronously detect changes in pixel brightness with high temporal resolution, high dynamic range, and low-power, have been very successful for robotic vision and autonomous navigation. Depth estimation from monocular or stereo event cameras is an important task which helps improve downstream robotic localization and mapping...

This algorithm is developed in the field of control systems for control of affine nonlinear continuous time (CT) systems, which include important application areas such as robotics and autonomous systems, aerospace, manufacturing, and process control. This invention falls under the class of continuous-time reinforcement learning (CT-RL) control algorithms....

Modern AI for strategic games including chess has evolved from early rule-based systems such as brute force engines that relied on handcrafted heuristics and exhaustive search, to sophisticated learning-based methods that can create and refine their own strategies. Reinforcement learning (RL) programs including Alpha Zero mastered chess starting with...