Digital Commons Network^™

Amorphous Form Of Aficamten And Its Preparation Process, Anonymous

Defensive Publications Series

This publication relates to amorphous form of Aficamten (I) and process for its preparation.

User-Driven Ai Model Context Restoration Based On Message Indexing, Aakash Phoughat

Defensive Publications Series

In long conversations, a chatbot powered by a large language model (LLM) may start to exhibit inaccurate responses or hallucinations due to context saturation. To continue the conversation, the user needs to identify the point of likely errors and initiate a separate conversation that includes the appropriate context by copy-pasting content or other manual actions. This disclosure presents techniques for context restoration with message indexing to address this problem. Per techniques described herein, conversation turns between a user and an AI model accessed via a chatbot interface are indexed and message indexes are displayed to the user. If the output …

Automated Project Initiation And Management Using Generative Ai And Organizational Knowledge Retrieval, Katherine Brown Armstrong, Kevin Woods, Magdalena Flores

Defensive Publications Series

Significant administrative overhead is often encountered in operational environments due to fragmented, manual processes for tracking diverse work types and managing project lifecycles. Reliance on individual effort across disparate applications leads to inconsistent execution and limited visibility into workflows. To address these limitations, a method is disclosed for context-aware workflow automation utilizing generative artificial intelligence. Techniques are grounded in a curated organizational knowledge base containing standard operating procedures and best practices. Digital interactions are analyzed to derive project context, triggering automated cross-application actions such as drafting project artifacts, identifying stakeholders, and populating tracking tools. Administrative burdens are reduced by automating …

Testing Calibrated Delay Circuit Propagation Delay With A Phase Frequency Detector, Karthikeyan Subramanian, Kasi Chunduri, Mukesh Goyal, Dhivyabharathi Vethanayagam, Gurumoorthy, Durgesh Gupta

Defensive Publications Series

Testing calibrated delay circuit (CDC) propagation delay at high double data rate (DDR) speeds is often limited by inaccuracies and increased test time in automated test equipment (ATE). Traditional characterization methods may be impacted by external hardware parasitics that degrade accuracy, as well as specialized tools that may increase both test time and overall cost.

This disclosure describes an on-chip analog design-for-test (DFT) method that utilizes a phase frequency detector (PFD) from an existing phase-locked loop (PLL) to measure CDC delay. The calibrated clock and reference clock are routed to the PFD to convert the delay into a duty cycle. …

Method For Smart Left Turn And Merging With Powertrain Aware, Anonymous

Defensive Publications Series

Drivers frequently encounter significant challenges when making left turns onto main roads or merging into traffic, primarily due to difficulties in accurately assessing gaps in oncoming traffic. A critical concern arises from the misperception of approaching vehicle speeds and distances, a problem greatly exacerbated by the increasing prevalence of Electric Vehicles (EVs) and other quick-accelerating vehicles.

EVs deliver maximum torque instantaneously, allowing many to accelerate from 0 to 60 mph in under 5.0 seconds. This unique capability results in different acceleration characteristics compared to Internal Combustion Engine (ICE) vehicles, with some studies noting that EVs may accelerate more quickly. As …

Marker-Based Imperative Extraction Facility For Json Configurations, Anonymous

Defensive Publications Series

This disclosure describes a technique for extracting sub-JSONs from large JSON configurations (e.g., greater than 80,000 lines) without requiring JSON Schema definitions or knowledge of the JSON structure. The approach employs embedded marker annotations within JSON objects that specify filter criteria using a domain-specific language (DSL). A recursive parser evaluates these markers against user-selected filters to produce tailored configuration subsets on demand.

The DSL supports OR logic (comma-separated filters), AND logic (plus-sign conjunction), mixed boolean expressions, and recursive nested item evaluation via a checkNestedItems parameter. A key characteristic is imperative selection: elements without markers are included automatically in the output, …

Uart Virtualization, Debug And Trace In A Multi-Processor System, Mukul Mehta, Gopi Tummala, Prasad Avirneni

Defensive Publications Series

Accessing the UART (universal asynchronous receiver/transmitter) ports of a particular processor within a system-on-chip (SoC) generally requires exposing the UART pins of the processor at the die periphery. However, the die periphery has space only for a limited number of pins. This disclosure describes techniques that leverage universal serial bus type-C (USB-C) to communicate with processors over UART to enable frictionless SoC debugging. The techniques enable efficient use of SoC pins, rapid and parallel access to processor UARTs, and rapid root-cause analysis. On devices where USB is the only exposed interface, debugging can be facilitated by simply connecting the USB …

A System And Method For Real-Time Artificial Intelligence Driven Image Personalization For Digital Advertisements, Tarushi Dubey, Sumit Kumar, Pradeep Choube, Shivam Mohan, Vinay Sarda, Ravi Shanker Kumar Sinha

Defensive Publications Series

The present disclosure provides a method and a system for real-time artificial intelligence driven image personalization for digital advertisements. The method includes collecting the user data comprising user behavior data, and user transactions data and generating a user preference vector and a personalization sensitivity vector based on the user data. The method further includes determining a base advertising image comprising editable and locked components. The method includes determining personalized editable components based on the personalization sensitivity vector and generating a personalized advertising image based on the editing of the personalized editing components, user preference vectors and brand guidelines. The method …

System And Method For Preventing Insider Fraud Using Change-And-Revert Pattern Detection In Sensitive Banking Fields, Alok Roy

Defensive Publications Series

The present disclosure relates to a technique for preventing insider fraud using change-and-revert pattern detection in sensitive banking fields. The technique includes a monitoring framework configured to detect alterations to protected customer identity fields, such as mobile numbers and email addresses, recovery contact, card details, communication preference, or authentication identifiers. An audit trigger logs modification events, capturing details such as old and new values, timestamps, and user identifiers. A change data capture service continuously polls these logs to identify changes and activate cooling-off periods and transaction restrictions upon detection. The policy enforcement module invokes these restrictions, while the fraud pattern …

Dynamic Alpha Channel Tone Mapping And Multi-Factor Transparency Modulation, Yannis Guyon, Vincent Rabaud

Defensive Publications Series

Dynamic alpha tone mapping optimizes transparency for non-opaque digital content across varying environmental and digital contexts. An environmental sensing module, a content intent analysis module, and a composition analysis module concurrently ingest ambient luminous intensity, semantic metadata, and spatial frequency measurements of underlying content layers. A mathematical interpolation engine normalizes the data into a unified coefficient scale to select a non-linear Alpha Transformation Function. The Alpha Transformation Function, implemented as a localized gain map or a global gain curve, modifies the alpha channel of graphical assets during rendering. The Alpha Transformation Function facilitates increased contrast and legibility for high-priority assets …

Fusing Fitness And Health Sensor Data From Multiple Sources, Justin Phillips, Daniel Roggen, Robert Harle

Defensive Publications Series

Estimates of cardiorespiratory fitness metrics (e.g., maximal oxygen uptake) from wearable electronic devices are often characterized by significant noise and variability. Inconsistencies arise due to algorithmic differences between manufacturers, sensor inaccuracies, and biological variance. The integration of data from multiple disparate sources lacks a standardized method for weighting measurements based on source reliability or handling late-arriving data. While the disclosed method is described in the context of cardiorespiratory fitness metrics such as VO2 max, it is equally applicable to any changing physiological variable derived from noisy or disparate sources, including but not limited to daily or weekly average blood pressure. …

Automated Detection And Identification Of Content Sources Using Invisible Watermarking, Anonymous

Defensive Publications Series

Systems, methods, and computer-readable media are described for automated detection and identification of content sources using invisible watermarking.

Distributed Fifo On-Demand Queuing Container Lifecycle Orchestration System For Heavy Workloads, Chang Kang Charles Goh, Hitika Mahapatra

Defensive Publications Series

The present disclosure relates to distributed container orchestration and, more particularly, to a First-In-First-Out (FIFO) on-demand queuing container lifecycle orchestration system for executing heavy workloads. The disclosed system receives on-demand workload requests and queues the requests based on submission time for processing by a distributed worker pool. Each worker node is restricted to executing a single workload at a time and acquires execution locks to prevent concurrent container execution conflicts. Upon selection, a worker node dynamically creates a container to execute the workload, records execution results and operational logs, and destroys the container upon completion to release computational resources. A …

Deterministic Application Install Attribution Via Custom Store Page Parameterization, Ishan Bansal, Meng He, Jenn Park

Defensive Publications Series

The present disclosure relates to systems and methods for attributing application installations to digital advertising interactions within identifier-restricted environments. The system utilizes customizable application store landing pages to transport static attribution parameters from an advertising campaign to an installed application. An advertising entity generates a static parameter associated with a campaign, which is subsequently embedded into the configuration of a custom product page on a digital application distribution platform. When a user navigates from an advertisement to this custom page and actuates a specific interface element (such as an “Open” button) immediately following installation, the application store platform passes the …

Proactive Learned Indexing Driven By Query Workload Forecasting, Mukesh Kumar Marodia, Virender Kumar Singla, Saurabh Uttam

Defensive Publications Series

Systems for data indexing can involve a trade-off between the upfront cost of static indexes and potentially suboptimal performance from on-demand data scanning, while some reactive adaptive methods may incur an initial query latency penalty. This disclosure describes systems and methods for predictive, learned adaptive indexing. A forecasting component, which may use a sequence-aware machine learning model, can analyze historical query patterns to predict future data access ranges. Based on these predictions, a system can speculatively trigger just-in-time index construction on anticipated data partitions, for example, using available system resources. This proactive approach, which can be combined with a reactive …

Cloud Cost Modeling Using Architectural Patterns And Heuristic Translation, Thiyagaraj Krishna

Defensive Publications Series

Forecasting cloud infrastructure costs can present challenges due to the difficulty in translating high-level business objectives, such as projected user volume, into the granular technical inputs that some estimation tools may be configured to use. Systems and methods are described for cloud cost modeling that can utilize a repository of predefined architectural patterns and associated heuristic rules. A system can receive business-centric inputs, such as application type and volume metrics, select a corresponding architectural pattern, and apply heuristic logic to translate these inputs into a manifest of technical resource demands, for example, compute, storage, and networking resources. This process may …

Creating Realistic Virtual Spaces Using Laser Scans And Panoramic Images, Qian Zhang, Yunwen Zhou, Erin Hong, Eric Lee Turner, Dinghuang Ji, Jürgen Sturm

Defensive Publications Series

A method is disclosed for generating scalable, high-fidelity scene replicas by optimizing 3D Gaussian Splats directly within the equirectangular domain. Conventional pipelines often convert spherical scanner data into perspective views or surface meshes, a process that frequently introduces stitching artifacts, geometric seams, and floaters. To address these limitations, the described approach utilizes a differentiable panorama splat rasterizer. This component processes registered laser scan data and 3D point clouds natively in a spherical format, without projecting imagery onto flat perspective planes. By refining spherical harmonic coefficients and opacity through backpropagation against the original panoramic inputs, the method effectively mitigates conversion-induced distortions. …

Techniques For Qr-Enabled Card Capture For Mobile Payments, Dileep R. Dominic

Defensive Publications Series

The present disclosure relates to a method for facilitating secure card-not-present (CNP) mobile payments through the use of a physical payment card embedded with an encrypted QR code. The QR code encodes essential cardholder data, including a primary account number (PAN) and expiry date. The method involves recognizing a designated input field on a mobile device while accessing a payment gateway interface. An interface module displays a camera icon adjacent to the input field, enabling a scanning procedure that utilizes a native camera application to read the QR code. This scanning process automatically populates the input field with the decoded …

A Method And Payment System For Preventing Card Skimming During Payment Transactions, Binayak Biswas

Defensive Publications Series

The present disclosure relates to a method and payment system for preventing card skimming during payment transactions. The method involves generating a dynamic PIN token using a secret key and transaction-specific data. The dynamic PIN token is transmitted via a near-field communication (NFC) interface from a mobile device or contactless card to a point-of-sale (POS) terminal. This token is verified at a verification network utilizing the same secret key and transaction-specific data to confirm the authenticity of the payment transaction. Upon successful verification of the dynamic PIN token, the payment transaction is executed, ensuring that the token is unique for …

Self-Launching Boat (Slb), Colin Hilton

Defensive Publications Series

A maritime drone able to operate in displacement and planing modes on water, or as a sled upon surfaces like snow, ice or sandbanks. An ancillary feature of the craft is that as few as four motors allow it to be launched from land to water in the absence of any facilities, and afterward fly within 'surface-effect' should conditions permit.

Innovative Labyrinth Seal Design To Optimize Leakage And Thrust Loads In Refrigerant Compressor

Defensive Publications Series

Labyrinth seals play a crucial role in shaft sealing applications in protecting bearings and preventing contamination. However, they tend to have high leakage values which negatively impact compressor efficiency. This proposed solution of labyrinth seal provides an axial and radial arrangement of labyrinth design that reduces the leakage drastically and also helps in reducing the effective thrust load of the compressor.

Turbine Expander Wheel With An Integrated Thrust Generator

Defensive Publications Series

Designing a multi-stage centrifugal compressor involves several challenges, particularly concerning thrust and clearances. Efficient thrust balancing is essential to prevent excessive wear on bearings and ensure stable operation while optimizing tip clearance is essential to avoid performance losses. The proposed solution discloses a radial turbine wheel with a radially extended back disk that acts as a counterbalance thrust generator.

Reversed Compressor Housing Ported Shroud Rib

Defensive Publications Series

Ported Shroud for a turbocharger allows air from downstream of compressor to recirculate back to upstream to increase flow stability and avoid surge. To expand the operating range of a centrifugal compressor, shroud walls often have annular ribs. This invention proposes a reversed design of ported shroud rib that improves compressor surge behavior and widens compressor map without any additional cost.

Dynamic Channel Selection Via Persistent Client-Side State Mapping, Abhishek Prasad

Defensive Publications Series

Manual channel selection can be required when a user signs into a primary account with multiple corresponding profiles on a content sharing platform, provided that no default profile is configured. This results in redundant user prompts and inefficient use of network resources through additional round-trip requests. This disclosure relates to dynamically selecting an active channel based on historical activity on a specific client device. The identity of the last active channel used on the client device can be stored using authentication data (e.g., an encrypted state cookie). When a subsequent sign-in is initiated, the state cookie can be retrieved and …

Multimodal Retrieval-Augmented Generation For Context-Aware Source Code Modification, Zheng Liang, Aileme Omogbai, Erin Altenhof

Defensive Publications Series

Generative artificial intelligence models may face challenges when modifying components within large software systems, as they can lack the project-specific context to link a visual user interface element to its corresponding source code files. A system can utilize a multimodal retrieval-augmented generation approach where historical code revisions, which may include associated user interface images and textual descriptions, are processed with an embedding model to create a searchable vector index. At execution time, a user query, which can comprise a screenshot and a text prompt, may be used to search the index and retrieve semantically similar historical changes. This retrieved information, …

Scalable Personalized Score Approximation Via Core Subgraph Pre-Computation, Leonid Kuligin

Defensive Publications Series

Computing link analysis algorithm scores for individual nodes in a large-scale graph can be computationally expensive, which may limit practical application. A technique for approximating such a computation may be applied to graphs that can be partitioned into a core subgraph (G1) and a peripheral subgraph (G2) without edges pointing from G1 to G2. The approach may involve pre-computing and caching pruned vectors for nodes within the G1 subgraph. A link analysis algorithm score for a node in the G2 subgraph can then be approximated by calculating a linear combination, such as an arithmetic mean, of the cached vectors corresponding …

Adaptive Correction Of Digital Map Features Using A Text-Guided Generative Model, Yotam Intrator, Idan Kligvasser, Ehud Rivlin, Amir Livne

Defensive Publications Series

Static machine learning models applied in automated cartography may produce inaccurate map features or may not adapt to localized guidelines without potentially costly retraining. A post-processing framework can utilize a multi-modal, text-guided generative model to perform adaptive corrections. For example, a system can receive a visual representation of a map segment, such as a raster image of road markings, and a set of natural language guidelines as input. The generative model can then process these inputs to synthesize a new visual representation where map features may be altered to conform with the specified guidelines. This approach can decouple correction logic …

Automated Generation Of Product Compliance Briefs Using A Generative Model And Enterprise Data Sources, Venkat Sharma Gaddala, Gary Borella, Krish Mohan, Nanda Balasubramanian, Raghu Ravisankar, Shaswat Kumar, Joe Heinrich, Debbie Masada

Defensive Publications Series

The manual creation of specialized compliance documentation for international trade may be a resource-intensive process, potentially leading to operational delays. Systems and methods are described for the automated generation of product briefing documents. The technology may utilize a central orchestration service that integrates a generative artificial intelligence model with multiple enterprise data sources, for example, product lifecycle management systems and repositories of existing documents. To generate a new document, the system can identify similar products based on shared attributes, such as a commodity code, and could use data from these products to construct a few-shot prompt for the model. This …

Cascaded Model Architecture For Resource-Efficient Contextual Action Suggestion, Lixia Liu, Jian Wu, Georgi Angelov, Jiang Wang

Defensive Publications Series

Providing real-time contextual action suggestions in communication environments can involve computational cost and latency associated with running large-scale artificial intelligence models. A system can address these considerations using a cascaded, multi-layered modeling architecture. The system may first process contextual data, such as live meeting transcripts, through a lightweight, computationally inexpensive classification model. If this model returns a prediction with a confidence score below a predetermined threshold, the request can be escalated to a more resource-intensive model, such as a large language model, for a more nuanced analysis. This gating mechanism can balance predictive accuracy with computational efficiency, which may reduce …

Dynamic Distillation Cache For Augmenting A Local Generative Model With Teacher Model Reasoning, Ben Mccormack, Pablo Rodriguez, Karin Breitman, Sol Chea, Orna Berry Keren

Defensive Publications Series

Organizations in regulated or disconnected environments may face challenges in deploying high-performance generative models due to data residency or connectivity constraints, while local models can exhibit a performance gap for certain complex tasks. A hybrid architecture can augment a local student generative model using a dynamic distillation cache. This system can capture outputs, such as final answers, and underlying reasoning patterns from a remote teacher model. When a new query is received, the system can use semantic similarity to retrieve relevant cached reasoning and provide it as context to the local student model. This method of in-context distillation may allow …