Abstract: The disclosed computer-implemented method may include accessing various antenna elements and identifying parameters for an antenna that is to be formed using the accessed antenna elements. The method may also include assembling the antenna elements, using an artificial intelligence (AI) instance, into an assembled antenna that at least partially complies with the identified parameters. Various other methods, systems, and computer-readable media are also disclosed.

Abstract: A disclosed computer-implemented method may include (1) receiving (A) a set of design specifications associated with an antenna performance characteristic, (B) a set of requirements for an antenna architecture, (C) a parameterization that describes parameters of an antenna structure, and (D) a set of bounds for the parameterization, and (2) determining, based on the set of design specifications, the set of requirements, the parameterization, and the set of bounds, and in accordance with a global optimization algorithm and a local tuning algorithm, a design for the antenna architecture. Various other methods, systems, and computer-readable media are also disclosed.

Abstract: The disclosed computer-implemented method may include generating, using a machine-learning model of a computing device, a set of antenna designs. The method may also include tokenizing, by the computing device, each antenna design in the generated set of antenna designs. Additionally, the method may include predicting, by the machine-learning model of the computing device, a frequency response for each tokenized antenna design. Furthermore, the method may include comparing, by the computing device, the frequency response for each tokenized antenna design. Finally, the method may include selecting, by the computing device based on the comparison, an antenna design that meets a performance threshold for the frequency response. Various other methods, systems, and computer-readable media are also disclosed.

Abstract: The disclosed computer-implemented method may include accessing various antenna elements and identifying parameters for an antenna that is to be formed using the accessed antenna elements. The method may also include assembling the antenna elements, using an artificial intelligence (AI) instance, into an assembled antenna that at least partially complies with the identified parameters. Various other methods, systems, and computer-readable media are also disclosed.

Abstract: Computer implemented systems are described that implement a differentiable neural architecture search (DNAS) engine executing on one or more processors. The DNAS engine is configured with a stochastic super net defining a layer-wise search space having a plurality of candidate layers, each of the candidate layers specifying one or more operators for a neural network architecture. Further, the DNAS engine is configured to process training data to train weights for the operators in the stochastic super net based on a loss function representing a latency of the respective operator on a target platform, and to select a set of candidate neural network architectures from the trained stochastic super net. The DNAS engine may, for example, be configured to train the stochastic super net by traversing the layer-wise search space using gradient-based optimization of network architecture distribution.

Abstract: Systems, methods, and non-transitory computer-readable media can train a sequence model to output respective captions, or portions of captions, for content items. A determination can be made that a user of the social networking system is posting a content item for publication through a social networking system. A set of captions, or portions of captions, can be determined for the content item being posted based at least in part on the sequence model. The set of captions, or portions of captions, can be provided as suggestions to the user for use in a caption describing the content item being posted.

Abstract: Techniques are disclosed for implementing an extensible, light-weight, flexible (ELF) processing platform that can efficiently capture state information from multiple threads during execution of instructions (e.g., an instance of a game). The ELF processing platform supports execution of multiple threads in a single process for parallel execution of multiple instances of the same or different program code or games. Upon capturing the state information, one or more threads may be executed in the ELF platform to compute one or more actions to perform at any state of execution by each of those threads. The threads can easily access the state information from a shared memory space and use the state information to implement rule-based and/or learning-based techniques for determining subsequent actions for execution for the threads.

Abstract: Techniques are disclosed for implementing an extensible, light-weight, flexible (ELF) processing platform that can efficiently capture state information from multiple threads during execution of instructions (e.g., an instance of a game). The ELF processing platform supports execution of multiple threads in a single process for parallel execution of multiple instances of the same or different program code or games. Upon capturing the state information, one or more threads may be executed in the ELF platform to compute one or more actions to perform at any state of execution by each of those threads. The threads can easily access the state information from a shared memory space and use the state information to implement rule-based and/or learning-based techniques for determining subsequent actions for execution for the threads.

Abstract: In one embodiment, a method for sending notifications includes accessing a data set, with each data being associated with a first user and including (1) a first action, first state, and first reward that are associated with a first time period, and (2) a second action and second state that are associated with a second time period. The data set may be used to train a recurrent machine-learning model configured to take as inputs a state and action associated with a user and a time period, and predict a cumulative reward over multiple time periods. The cumulative reward may be recurrently defined based on an application of the recurrent machine-learning model to the user's subsequent state and action. The trained model may be used to predict whether sending a notification to a target user is justified based on a predicted cumulative reward.

Abstract: Systems, methods, and non-transitory computer-readable media can train a sequence model to output respective captions, or portions of captions, for content items. A determination can be made that a user of the social networking system is posting a content item for publication through a social networking system. A set of captions, or portions of captions, can be determined for the content item being posted based at least in part on the sequence model. The set of captions, or portions of captions, can be provided as suggestions to the user for use in a caption describing the content item being posted.

Abstract: A system and method is provided for detecting and responding to emergency vehicles. In one aspect, one or more computing devices may identify a set of light sources from an image based at least in part on one or more templates, and may filter the set of light sources in order to identify one or more light sources corresponding to a potential emergency vehicle. Moreover, the one or more computing devices may determine whether any of the one or more light sources is flashing and whether any of the one or more light sources is associated with a particular type of the potential emergency vehicle. Further, the one or more computing devices may maneuver a vehicle based on the determination to yield in response to at least one of the one or more flashing light sources and the particular type of the emergency vehicle.