Abstract: Techniques for multi-output model forecasting are provided herein. An example method can include a computing system receiving a request to forecast a value for a variable at a future time point based upon a time series, the time series comprising a sequence of data points, each data point in the sequence of data points identifying a time point and at least one value associated with the time point. The computing system can predict, using a first trained machine learning model and based upon the times series, a plurality of forecast values for the future time point, the plurality of forecast values including: a first forecast value forecasted for the variable at the future time point; and a set of one or more forecast attribute values for one or more attributes of the time series, each of the set of one or more forecast attribute values predicted for the future time point.

Abstract: Systems and methods are provided for proctoring online exams that collects reference samples of a student's speech and video of them looking at their computer screen. Audio and video of the student are then captured during the exam. That audio and video is segmented and compared to the reference samples in order to identify segments that potentially contain at least one of another speaker who is not the student, the student looking away from the computer screen, another face that is not the student's. The frequency and severity of flagged segments are used to indicate the overall level of suspicion to an exam proctor.

Abstract: An example mobile computing device includes: a wireless communications interface configured to connect to a network deployed by a plurality of base stations; a controller for the wireless communications interface, the controller configured to: initiate a scan for foreign base stations; when the scan is complete, return to a home base station and initiate a home channel dwell time; receive a data packet from the home base station during the home channel dwell time; receive an indicator that transmission of buffered data packets from home base station is complete; and in response to the indicator, terminate the home channel dwell time.

Abstract: Various embodiments are directed to processing and executing digital asset transfers between two users. An example method includes receiving a data object defining a digital asset transfer, and the data object may be a query from the first user to initiate an outbound transfer to a second user or an approval of a request for digital asset units from the second user. The method further includes retrieving account balance data objects associated with the digital asset accounts for the first and second users. The method further includes, responsive to determining that the digital asset transfer satisfies one or more transfer conditions, executing the digital asset transfer by executing a closed-loop debit from the first user and executing a closed-loop credit to the second user. The method further includes updating the account balance data object to reflect the resulting balances of each digital asset account.

Abstract: Techniques for lock state based on a trigger condition are described and may be implemented to transition a mobile device to a locked state based on various conditions. For instance, a mobile device receives transmission of cloud content and transmits the cloud content to a connected remote display device for output. A trigger condition occurs based on the transmission of the cloud content, such as a decrease in data transmission related to the cloud content. The mobile device can transition to a locked state based at least in part on the trigger condition.

Abstract: Systems and methods are disclosed for detecting hallucinations in large language models (LLMs). An example method includes receiving a first prompt for submission to the first LLM, generating, using the first LLM, a plurality of semantically equivalent prompts to the first prompt, generating, using the first LLM, a first response to the first prompt and a plurality of second responses to the plurality of semantically equivalent prompts, generating, using a second LLM, a plurality of third responses to the semantically equivalent prompts, generating a semantic consistency score for the first response based at least in part on the first prompt, the plurality of semantically equivalent prompts, the plurality of second responses, and the plurality of third responses, and determining whether or not the first response is an accurate response to the first prompt based at least in part on the semantic consistency score.

Abstract: A first large language model (LLM) instance may be instructed to request data while being prevented from performing calculations using the data. A second LLM instance may be instructed to provide a response to the request for data based on a known complete data set. The response may be translated into a machine-readable response in a format configured for processing by a calculation engine. The calculation engine may process the machine-readable response, thereby generating a calculation engine output. A mismatch between the calculation engine output and a known result obtained using the known complete data set may be identified, and the instruction to the first LLM may be modified in response.

Abstract: A method for generating content using a diffusion model of an electronic device, may include: obtaining latent vectors of an input content; inputting the latent vectors into a first lightweight adapter configured for the first application type from among a plurality of lightweight adapters configured individually for application types of the plurality of applications; transforming the latent vectors of the input content into a plurality of intermediate latent vectors using the first lightweight adapter; performing a denoising operation to transform the plurality of intermediate latent vectors into a plurality of next operation vectors; and generating the final content belonging to the application type by decoding the next operation vectors; and outputting the final content.

Abstract: Building machine learning models using a standardized library of machine learning model tools. An Application Programming Interface (API) serves as an interface between one or more computer applications that receive user commands for building machine learning models and the standardized library of tools. The API provides an interface between data scientists tasked with designing machine learning models conceptually and a standardized set of software engineering tools. The API enables incorporation of the relevant standardized software engineering tools to build the machine learning models that have been designed conceptually by the data scientists.

Abstract: A method for operating a hybrid energy network comprising a plurality of power sources is described. After receiving information indicating predicted operating conditions of the hybrid energy network, the values of one or more constraints associated with an objective function that defines a relationship between a plurality of performance parameters associated with the hybrid energy network may be updated. Using a quantum computing system, an optimal value of each of the plurality of performance parameters may be determined. Based on the optimal values of each of the plurality of performance parameters, an operating plan for the hybrid energy network may be determined. Adjustments may be made to at least one of power sources according to the generated operating plan.

Abstract: A package comprising a substrate, an integrated device coupled to a first surface of the substrate through at least a first plurality of solder interconnects; and an antenna device coupled to a second surface of the substrate through at least a second plurality of solder interconnects. The antenna device comprises an antenna device dielectric layer; at least one antenna located on a first surface of the antenna device dielectric layer; a shield located on at least one lateral surface of the antenna device; and a slot opening on the first surface of the antenna device.

Abstract: A system includes a VCSEL element, a dielectric waveguide, and a dielectric grating coupler. The VCSEL element includes a first reflector region, a second reflector region opposite to the first reflector region, and an active region between the first reflector region and second reflector region. The dielectric waveguide is integrated with the VCSEL element. The grating coupler is formed on the dielectric waveguide for coupling an electromagnetic wave of the VCSEL element into the dielectric waveguide.

Abstract: A method of providing media content to a communication device. The method comprises retrieving subscriber attributes by a media content provider application from a subscriber database; creating a manifest of media content artifacts by the media content provider application based on identifying alignment between at least some of the media content artifacts, attributes of the communication device, and the subscriber attributes, wherein the manifest of media content artifacts identifies a list of selected media content artifacts along with a location from which the communication device can retrieve the selected media content artifacts and identifies a retrieval priority of each of the selected media content artifacts; and sending the manifest by the media content provider application to the communication device, whereby the communication device is enabled to retrieve the selected media content artifacts in a prioritized manner and install some of the selected media content artifacts in the communication device.

Abstract: Methods and apparatus configured to obtain a decision-point value, and send, for a logical channel group (LCG) having a quantity of data pending an uplink transmission, a long buffer status report in response to the decision-point value exceeding a threshold value, or a short buffer status report in response to the decision-point value being equal to or less than the threshold value are disclosed. The decision-point value may be a buffer status report-type determinative value, which may be based on a peak power envelope of a wireless communication device, a data transmission rate historically obtained by the wireless communication device, a number of component carriers available for a complete upload of a buffer holding data, a cost function, an amount of data associated with the LCG that is pending the uplink transmission, a type of the wireless communication device, or a latency of communications of the wireless communication device.

Abstract: A computer implemented method for maintaining information security infrastructure of an entity, includes accessing a cascading classification schema comprising a plurality of levels of classifications indicative of critical security controls, identifying a plurality of Key Risk Indicators (KRIs) and assigning a plurality of respective KRI weight values, a plurality of respective first KRI threshold values, and a plurality of respective second KRI threshold values to the plurality of respective KRIs, assigning the plurality of KRIs to the lowest level of classification, defining a plurality of data points corresponding to the plurality of KRIs, determining a plurality of KRI score values corresponding to the plurality of respective KRIs by comparing a plurality of data values associated with the plurality of respective data points, with the plurality of first KRI threshold values and the plurality of second KRI threshold values, and determining a cascading schema of aggregate scores.

Abstract: Fiber optic networks, optical network units, and methods of configuring a fiber optic network and an optical network unit using a variable ratio coupler are disclosed. In one embodiment, a fiber optic network includes a feeder optical cable. The first optical network unit and the second optical network unit each includes a variable ratio coupler operable to selectively set a split-ratio among a plurality of split-ratios, and a fiber optic transceiver module. The fiber optic network further includes a first drop cable optically coupling the first optical network unit to the feeder optical cable at a first position, and a second drop cable optically coupling the second optical network unit to the feeder optical cable at a second position. A first split-ratio of the first optical network unit is based on the first position and a second split-ratio of the second optical network unit is based on the second position.

Abstract: A user equipment (UE) may receive a downlink message from a non-terrestrial network (NTN) which may include one or more sets of terrestrial network (TN) cell reselection information. The one or more sets of TN cell reselection information may further include location information corresponding to one or more cells of the TN. The UE may determine its location and, in response to receiving the downlink message, compare its location to the location information corresponding to the one or more cells of the TN. The UE may then select, based on the comparison, a set of the one or more sets of TN cell reselection information and further perform a cell reselection procedure based on the selection.

Abstract: Various aspects of the present disclosure generally relate to wireless communication. In some aspects, a user equipment (UE) may obtain an indication to perform a selective secondary cell group activation or a subsequent conditional handover associated with a subsequent conditional primary cell change, the subsequent conditional handover associated with the subsequent conditional primary cell change being performed after a previous conditional handover associated with a previous conditional primary cell change. The UE may initiate the selective secondary cell group activation or the subsequent conditional handover. The UE may detect a failure occurrence associated with the selective secondary cell group activation or the subsequent conditional handover.

Abstract: An apparatus configured to process, based on signaling received from a child device, a data request, determine whether an active context exists with the child device based on data stored in firmware and determine whether to respond to the data request based on whether the active context exists with the child device.

Abstract: This disclosure generally relate to a method and system for network policy simulation in a distributed computing system. The present technology relates techniques that enable simulation of a new network policy with regard to its effects on the network data flow. By enabling a simulation data flow that is parallel and independent from the regular data flow, the present technology can provide optimized network security management with improved efficiency.