Abstract: Systems and methods are described for predicting tasks that users perform on websites. More specifically, a machine learning model is trained using page session data (that includes page flows which are a sequence of pages visited by a user) for each user that is associated with the website. In addition, the machine learning model is fine-tuned using previous page flows associated with the user, user attribute data that is associated with the user, and user tasks, which are tasks that have been previously completed by the user. When a user visits the website, the machine learning model determines a predicted task after the user visits a predetermined number of webpages. A recommended content is generated based on the predicted task and displayed before the user.

Abstract: Computer-implemented methods are directed to gap and compliance analysis for technical documents. Aspects include determining an assessment type for a technical document received from a user device. Aspects also include analyzing the technical document using a set of microguidelines associated with the assessment type. Aspects further include generating an alignment score using the set of microguidelines. Aspects also include generating a recommendation for the technical document based on the alignment score and the set of microguidelines. Aspects further include initiating a modification to the technical document based on the recommendation for the technical document.

Abstract: In various examples, an image search is performed based on a query image where the result includes images with semantic information that matches the query image. For example, image encoders are used to extract information from images that are stored in an index. Continuing this example, when a query is obtained the index is searched to determine potential matches, and a comparison is performed between the query image and the potential matches to determine images with semantic information matching the query image.

Abstract: In various examples, an image search is performed based on a query image where the result includes images with semantic information that matches the query image. For example, image encoders are used to extract information from images that are stored in an index. Continuing this example, when a query is obtained the index is searched to determine potential matches, and a comparison is performed between the query image and the potential matches to determine images with semantic information matching the query image.

Abstract: Embodiments relate to providing prompt recommendations for artificial intelligence (AI) engines. Aspects include determining that a received prompt is in a cluster, the cluster having a representative prompt and generating candidate prompts based on the representative prompt. Aspects include ranking the candidate prompts to determine a selected candidate prompt from the candidate prompts; and presenting the selected candidate prompt to be input for an AI engine.

Abstract: In some implementations, a network device may receive an indication of a next-next-hop of the network device and link quality information of the next-next-hop. The network device may transmit network traffic based on the indication of the next-next-hop and the link quality information of the next-next-hop.

Abstract: A system for providing enhanced search functionality is provided. A system may receive a prefix from a user device and identify that a user's intent is to do an exploratory query. The system may further identify a topic based on the prefix and generate one or more associated queries relating to the topic. The system may further generate one or more headings relating to the topic. The system may then provide one or more of the one or more headings and the one or more associated queries relating to the topic to a user device.

Abstract: Matrix multiplication operations can be implemented, at least in part, on one or more tensor cores of a parallel processing unit. An efficiency of the matrix multiplication operations can be improved in cases where one of the input operands or the output operand of the matrix multiplication operation is a square matrix having a triangular data pattern. In such cases, the number of computations performed by the tensor cores of the parallel processing unit can be reduced by dropping computations and/or masking out elements of the square matrix input operand on one side of the main diagonal of the square matrix. In other cases where the output operand exhibits the triangular data pattern, computations can be dropped or masked out for the invalid side of the main diagonal of the square matrix. In an embodiment, a library implementing the matrix multiplication operations is provided.

Abstract: Various examples are directed to systems and methods for managing multiple cloud-delivered services. An integration adapter may receive, from a first cloud-delivered service, event data describing an event at the first cloud-delivered service. The integration adapter may send a request for data objects for responding to the event to a determination engine executing at the data distribution computing system. The determination engine may determine a first data object and a second data object responsive to the request for data objects. The integration adapter may replicate the first data object from a database management system to generate a first replicated data object and replicate the second data object from a second cloud-delivered service to generate a second replicated data object. The integration adapter may send the first replicated data object and the second replicated data object to a target cloud-delivered service.

Abstract: Methods, apparatus, and processor-readable storage media for automated resource prioritization using artificial intelligence techniques are provided herein.

Abstract: Systems and methods for generating autocomplete text using a language model are disclosed. An image and text-prefix may be entered at an input field of a search application. The image is processed to generate an image description. The image description and the text-prefix signals may be used as input at a language model to generate an autocomplete text by the language model. A contextual history may also be included as input to the language model. The autocomplete text is an output by the language model based on the input at the language model. The auto-complete text may be a next-word ghosting.

Abstract: This disclosure describes a query gateway system that provides an efficient and flexible framework for providing context-retained autosuggest queries from an autosuggest query system (e.g., a search engine query experience) to a generative language model system (e.g., an AI chat experience). For instance, the query gateway system establishes a framework to leverage the features and services of the autosuggest query system and automatically provides context-retained queries to the generative language model system using separate user interfaces that do not disrupt user navigation or require manual duplicative user input. Additionally, the query gateway system incorporates additional enhancements, including an AI chat eligibility model and a query reformulation model, to improve the computational efficiency and accuracy of the AI chat system.

Abstract: Various aspects of a system and method to provide driving assistance to safely overtake a vehicle are disclosed herein. In accordance with an embodiment, an electronic control unit used in a first vehicle is configured to detect a second vehicle in front of the first vehicle. A first position associated with the first vehicle and a second position associated with the detected second vehicle is determined for a first time instance. It may be determined whether a lateral distance between the determined first position and the determined second position is below a pre-defined threshold distance. A first alert is generated when the determined lateral distance is below the pre-defined threshold distance.

Abstract: Examples described herein provide a method that includes receiving a set of air-gapped systems' states from a plurality of monitored air-gapped systems. The method further includes aggregating one or more subsets of the set of air-gapped systems' states based on a vectorized set of standard operating procedures in execution at a time of the generating. The method further includes determining, using at least one aggregation, an intersystem state event dependency defining the aggregating using natural language processing techniques. The method further includes forwarding the intersystem state event dependency to a secondary event aggregator. The method further includes classifying the intersystem state event dependency as an abnormal intersystem state event dependency. The method further includes determining, based on a historical corpus of intersystem state event dependencies, a resolution for the abnormal intersystem state event dependency.

Abstract: A system and method for application security profiling that includes extracting a code property graph from at least a subset of a code base; generating a code profile from the code property graph, wherein generating the code profile occurs prior to a compilation of the code base; and applying the code profile, comprising of identifying sections of interest within the code base.

Abstract: An optical cavity includes a cell having at least one structure with a first end and a second end, a first end-cap configured to removably receive a first mirror, and a second end-cap configured to removably receive a second mirror. The first end-cap is attached to the structure at the first end and has an opening with a center axis. The second end-cap is attached to the structure at the second end and has an opening with a center axis that is axially aligned with the center axis of the first end-cap. The cell can be a closed cell, wherein the structure is a tube axially aligned with the center axis. The cell can be an open cell, wherein the structure is a tube that includes one or more openings through its wall. The cell can be an open cell, wherein the structure includes one or more rods, each off axis from the center axis.

Abstract: Disclosed are various embodiments for an intelligent trip coordinator. A trip entry associated with a user can be received by a computing device, where the trip entry includes trip data. The computing device can generate a trip record based at least in part on the trip entry, where the trip record includes the trip data. Next, the computing device can send the trip record to an approval service. In response to receiving an approval from the approval service, the computing device can forward the trip record to a booking service. Finally, the computing device modifies the trip record based at least in part on booking data provided by the booking service.

Abstract: A building management system, such as a small or medium business, having one or more control devices. Assigning codes to devices, and reading and listing them may aid in configuring the devices. Templates may be used for dynamic configuration of devices and equipment. Representing terminal assignments and wiring diagrams of control devices may be intuitive in that they resemble real hardware for ease of complete installation. The present system and approach may provide an intuitive way of securely registering devices with cloud usage, for example, involving a mobile phone with no manual entry of data so that a user can complete this process with ease. An intuitive, automatic and asynchronous device configuration downloading may be seen, such as to multiple control devices in a single shot. There may be seamless configuration data synchronization, for example, in view of replacement devices, or so the latest data is readily available.

Abstract: Systems and methods for generating autocomplete text using a language model are disclosed. An image and text-prefix may be entered at an input field of a search application. The image is processed to generate an image description. The image description and the text-prefix signals may be used as input at a language model to generate an autocomplete text by the language model. A contextual history may also be included as input to the language model. The autocomplete text is an output by the language model based on the input at the language model. The auto-complete text may be a next-word ghosting.

Abstract: Various aspects of a system and method to provide driving assistance to safely overtake a vehicle are disclosed herein. In accordance with an embodiment, an electronic control unit used in a first vehicle is configured to detect a second vehicle in front of the first vehicle. A first position associated with the first vehicle and a second position associated with the detected second vehicle is determined for a first time instance. It may be determined whether a lateral distance between the determined first position and the determined second position is below a pre-defined threshold distance. A first alert is generated when the determined lateral distance is below the pre-defined threshold distance.