Abstract: The description relates to automated binary code summarization. In one example, a binary code summarization tool receives binary code and combines the received binary code with natural language in a prompt for a large language model (LLM). The binary code summarization tool receives a semantic summarization from the LLM relating to the received binary code and evaluates the new semantic summarization for malicious functionality in the received binary code.

Abstract: A data processing system implements techniques for generating personalized content using a brand kit. The system receives a natural language prompt to generate content in a design application on the client device of a user and analyzes the prompt to determine whether the user intends to apply a brand kit to the generated content. The system automatically generates a brand kit for the user if one does not already exist and applies the brand kit to content generated using one or more generative models to create personalized content. The system includes a prompt generation unit that generates a plurality of model-specific prompts to the one or more generative models to cause the one or more generative models to create the personalized content.

Abstract: Solutions for evaluating source code generators use offline and online evaluation stages. Offline evaluation includes separating each of a plurality of input passages of software code into a plurality of constituent blocks. Each code generator (of a plurality of code generators) generates an equivalent block corresponding to each constituent block. A coding score is determined for each equivalent block (for each code generator), and the coding scores are aggregated across the equivalent blocks to provide an aggregate score for each code generator. A ranking of the aggregate scores is used to down-select to a fewer number of code generators for online evaluation. For this stage, the code generators output passages of software code, and user acceptance of the code generators' outputs may be used for further ranking and down-selection. Some examples weight the coding score according to a code utility estimate of the constituent blocks for which equivalent blocks are generated.

Abstract: A computing device is provided, including processor and a storage device holding instructions that are executable by the processor to implement a base artificial intelligence (AI) model and two or more delta AI models, each delta AI model having lower dimensionality than the base AI model. An inference request including an input prompt is received, the inference request specifying a selected delta AI model of the two or more delta AI models. The input prompt is input to the base AI model to thereby generate a base model result vector. The input prompt is input to the selected delta AI model to thereby generate a delta model result vector. An output vector is generated by combining the base model result vector and the delta model result vector via a combination operation. The output vector is output.

Abstract: A computer-implemented method for selective indexing of target content is disclosed. A web hosting system hosting the target content can collect user access data for the target content, which is presented in a first language; extract user locations from the user access data; detect, from the first user locations, an area associated with a second language that is different from the first language; and evaluate a trigger condition based at least in part on comparing a content metric, which measures user access to the target content from the area, to a content threshold. Responsive to detecting satisfaction of the trigger condition, the system can translate the target content from the first language to the second language; and index the target content presented in the second language so as to enable the target content to be searched using the second language.

Abstract: A media server uses selective just-in-time (“JIT”) transcoding of media such as video. For example, the media server determines a measure of complexity of a given segment of a given media sequence. The given segment has been encoded at a base bit rate. The media server evaluates a complexity condition for the given segment. As part of evaluating the complexity condition, the media server compares the measure of complexity to a complexity threshold. Based at least in part on whether the complexity condition is satisfied, the media server selects between use of preemptive transcoding and use of JIT transcoding for the given segment at a given target bit rate. In this way, the media server can selectively incur the cost of preemptive transcoding operations for the given segment if JIT transcoding would likely introduce unacceptable delay, and the media server can otherwise use JIT transcoding operations for the given segment.

Abstract: A cache service provides applications in a containerized, multi-tenant cloud-computing system low-latency access to secrets. The cache service may operate as a cluster-level service or a sidecar service. The cache service may store copies of secrets (which are located in one or more absolute stores) in a cache storage. The cache service and the cache storage may be closer to the applications than the one or more absolute stores are to the applications. The cache service may aggregate secrets associated with multiple entities in a single cache storage. The cache service may support isolation between secrets such that secrets of a first entity are isolated from secrets of a second entity. The cache service may enforce granulated access controls such that it can apply different access controls to secrets of a first entity than to secrets of a second entity.

Abstract: The present disclosure relate to highlighting audience members with reactions to a presenter of an online meeting. Unlike physical, fact-to-face meeting that enables spontaneous interactions among the presenter and the audiences that are collocated with the presenter, presenting materials during an online meeting raises an issue of the present not being able to see real-time reactions or feedback by the audience members. The present disclosure addresses the issue by dynamically determining one or more audience members who indicate reactions during the online meeting or presentation and displaying faces of the one or more audience members under spotlight to the presenter. The presenter sees faces of the audience members with reactions during the online presentation and responds to the audience members and keep the audience engaged. The spotlight audience server analyzes video frames and determines types of reactions of the audience members.

Abstract: A supply chain tracking system utilizes tracking codes to track products through a supply chain. A tracking code is assigned to each product. If the product is grouped with other products at a stage in the supply chain, a tracking code is assigned to the group, and the tracking code for each of the products in the group is associated with the tracking code for the group. If the group of products is further aggregated with groups of other products, such as in a shipping container, a tracking code is assigned to the aggregated groups of products, and the tracking code for each of the groups of products is associated with the tracking code for the aggregated groups of products. The tracking codes are used to generate a supply chain graph which maps the travel of each product through the supply chain.

Abstract: Generally discussed herein are devices, systems, and methods for digital signature generation security. A method can include generating, by a first device, a first random number, in generating a signature for a communication, masking, using the first random number, only a private key, a hash of the communication, or a combination thereof, and providing the signature with the communication to a second device.

Abstract: A quantum measurement circuit implements a hypergraph product code (HPG). A syndrome can be extracted from the circuit by preparing a readout qubit of the quantum measurement circuit in a known state, preparing a row-based measurement gadget, and preparing a column-based measurement gadget in the quantum measurement circuit. The row-based measurement gadget entangles the readout qubit with a first subset of the target set of data qubits in a same row of the quantum measurement circuit as the readout qubit, and the column based gadget entangles the readout qubit with a second subset of the target set of data qubits in a same column of the quantum measurement circuit as the readout qubit. The syndrome is extracted by measuring the readout qubit to extract the parity of the target set of data qubits.

Abstract: Techniques for implementing a hybrid search for documents having a parent-child relationship are disclosed. In some embodiments, a computer-implemented method comprises: building a base index using a first table of parent documents and a second table of child documents, the base index storing document identifiers of the parent and child documents in contiguous document blocks, where each document block stores the document identifier of a parent documents and the document identifier of each one of the child documents that is related to the parent document in a sorted order; subsequent to the building of the base index, updating a live index to include document identifiers of additional parent documents that have been added the first table and document identifiers of additional child documents that have been added to the second table; and computing a hybrid set of search results for a search query using the base index and the live index.

Abstract: The present disclosure relates to systems, methods, and computer-readable media for securely verifying an identity of a user of a client device based on a signal transmitted by the client device. For example, systems disclosed herein include registering a client device and facility device via a cloud computing system to enable the client device and facility device to securely communicate a signal via a wireless connection. The systems disclosed herein additionally include determining whether a trigger condition applies based on a position of the client device relative to the facility device. The systems disclosed herein further include maintaining and updating a subset of user verification information to include personal verification of a registered user of the client device. Using the subset of user verification information, a biometric scanning device may efficiently and accurately verify an identity of an individual associated with the client device.

Abstract: Innovations in syntax and semantics of coded picture buffer removal delay (“CPBRD”) values potentially simplify splicing operations. For example, a video encoder sets a CPBRD value for a current picture that indicates an increment value relative to a nominal coded picture buffer removal time of a preceding picture in decoding order, regardless of whether the preceding picture has a buffering period SEI message. The encoder can signal the CPBRD value according to a single-value approach in which a flag indicates how to interpret the CPBRD value, according to a two-value approach in which another CPBRD value (having a different interpretation) is also signaled, or according to a two-value approach that uses a flag and a delta value. A corresponding video decoder receives and parses the CPBRD value for the current picture. A splicing tool can perform simple concatenation operations to splice bitstreams using CPBRD value for the current picture.

Abstract: Various embodiments discussed herein are directed to improving existing technologies by causing presentation, to one or more user devices associated with one or more meeting attendees, of one or more indications of one or more content items during or before a meeting based at least in part on one or more natural language utterances associated with the meeting, a context of the meeting, and/or a context associated with one or more meeting attendees. In other words, particular embodiments automatically recommend relevant content items responsive to the real-time natural language utterances in the meeting, and/or other context.

Abstract: Generally discussed herein are devices, systems, and methods for secure cryptographic masking. A method can include generating a first random number, determining a result of the first random number modulo a prime number resulting in a second random number, subtracting the second random number from the prime number resulting in a first subtraction result, adding a private key value to the first subtraction result resulting in a first split, and responsive to determining the private key value is less than the random number, providing the first split and the second random number as splits of the private key.

Abstract: In one example, a display unit comprises a display panel that is configured to display digital images. The display unit further comprises an at least partially transparent protective layer that is arranged above the display panel. The display unit further comprises a controller that is communicatively attached onto an upper surface of the display panel. A biometric sensor pattern is integrated in the controller, and the controller is configured to control the biometric sensor pattern.

Abstract: A computing system is provided, including a processor and memory executing a reboot tracking module configured to read out a stored reboot request identifier assigned to a node in the computing system including a first value, and receive a first reboot request to reboot the node in the computing system including a first reboot request identifier. The reboot tracking module is further configured to, responsive to identifying a match between a value of the first reboot request identifier and the first value of the stored reboot request identifier, accept the first reboot request and update the stored reboot request identifier with a second value, receive a second reboot request to reboot the node including a second reboot request identifier, and responsive to identifying a mismatch between a value of the second reboot request identifier and the second value of the stored reboot request identifier, reject the second reboot request.

Abstract: A method for text compression comprises recognizing a prefix string of one or more text characters preceding a target string of a plurality of text characters to be compressed. The prefix string is provided to a natural language generation (NLG) model configured to output one or more predicted continuations each having an associated rank. If the one or more predicted continuations include a matching predicted continuation relative to the next one or more text characters of the target string, the next one or more text characters are compressed as an NLG-type compressed representation. If no predicted continuations match the next one or more text characters of the target string, a longest matching entry in a compression dictionary is identified. The next one or more text characters of the target string are compressed as a dictionary-type compressed representation that includes the dictionary index value of the longest matching entry.

Abstract: One embodiment includes retrieving firewall flow log data that indicates whether a flow was allowed or denied, an identifier of a rule that allowed or denied the flow, a source port, a protocol, a destination port, a source IP or FQDN, and a destination IP or FQDN. The method continues with processing the firewall flow log data, such as by identifying and counting occurrences of unique flows and counting flows allowed or denied by each rule. The method further includes generating a recommendation of at least one of limiting an existing rule, deleting an existing rule, and modifying rule application precedence. The recommendation may be generated based on at least one of the occurrences of unique flows and counted flows allowed or denied by each rule of a rule base. This method also includes providing the recommendation within a user interface as a selectable option for implementation.