Abstract: A social networking system provides a user with a feed of content items associated with other users connected to the user via the social networking system. Additionally, the social networking system identifies additional content items for presentation to the user and generates an additional feed including the additional content items. The additional content items may be determined by identifying content items having various characteristics and scoring the content items based on the characteristics. Content items having at least a threshold score are identified as additional content items. Examples of characteristics of content items include users providing content items to the social networking system, locations associated with social networking system users, and interaction with content items by social networking system users. In some embodiments, the additional feed modifies presentation of the additional content items based on their associated scores.

Abstract: Exemplary embodiments relate to improvements in the design of a messaging inbox. The inbox may display different units or “modules” for providing a user with quick access to different inbox functionalities. Content may be ranked and ordered within a module, and modules may be ranked and ordered with respect to each other. Inter-module ranking may be based on elements such as individual activity and the activity of the messaging service's user base, the value of the module to the messaging service, and offline models. Intra-module ranking may be defined by individual modules (e.g., based on recency, importance to user, affinity, etc.). Intra-module ranking may affect inter-module ranking, such as when individual content in a low-rated module is particularly exciting or pertinent (causing the entire module to receive a higher inter-module ranking).

Abstract: A mechanism for thermal testing is described. A test vehicle includes a heating element, a thermal sensor and a processor. The processor is configured to control the heating element to output an amount of the energy per unit time. Temperature readings are received using the thermal sensor. A thermal property associated with a thermal mass is determined based at least in part the amount of the energy output and the received temperature readings.

Abstract: In one embodiment, a method includes, by a client system, receiving, by an assistant xbot of the client system, a request from a first user for a summary of user content from a first content source, retrieving, from the first content source, a plurality of content items corresponding to the request, generating a personalized summary of the retrieved content items, wherein the personalization of the summary is based on a user profile of the first user, and presenting, by the assistant xbot, the personalized summary responsive to the request within a separate communication interface between the assistant xbot and the first user, wherein the personalized summary is interactable by the first user to react to one or more of the plurality of content items.

Abstract: A computing device enhances digital content on a display based on a distance to an individual from the computing device. The computing device detects an individual in the environment around the computing device and determines a distance from the computing device to the individual. The computing device enhances the information included in the digital content when the distance is within a first threshold distance from the computing device. Similarly, the computing device removes information included in the digital content when the distance is outside of a second threshold distance from the computing device.

Abstract: A different combination of multiple portions of a content item is selected for display to each of multiple sets of users of an online messaging application. The different combination of portions of the content item is sent for display to the corresponding set of users within a user interface of the online messaging application in association with information identifying a set of message threads in which each user is participating. A request to view the entire content of the content item is then received from a subset of each of the sets of users and a performance metric associated with each combination of portions of the content item is tracked based on the received request. Based on the performance metric, a performance of each portion of the content item is evaluated and used to select an optimal combination of the portions of the content item.

Abstract: In one embodiment, a method includes receiving a user query associated with dialog-intents at a client system, executing tasks corresponding to the dialog-intents, generating a multi-perspective response by a stitching model based on two or more of execution results of the tasks, wherein the multi-perspective response comprises a natural-language response combining the two or more execution results, and presenting the multi-perspective response at the client system.

Abstract: An online system receives video data items from users and encodes the video data items using various codecs. To account for different computational resources used for encoding using different codecs, the online system ranks combinations of video data items by ratios of encoding video data items with different codecs to computational costs of encoding different video data items with different codecs. The benefit of encoding a video data item with a codec is based on a compression efficiency of the codec and a predicted aggregate amount of the video data item displayed to various users of the online system. Encoding video data items with codecs based on the determined ratios allows the online system to optimize a duration of video data having at least a threshold video quality to users.

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: In one embodiment, a method involves accessing training data, where the training data contains an ordered sequence of data associated with a plurality of entities, training one or more deep learning models to determine, from the ordered sequence of data, a first set of embeddings for each entity of the plurality of entities, where each entity has a plurality of entity attributes, determining, for each of the plurality of entity attributes, a corresponding initial embedding, training the one or more deep-learning models to refine the initial embeddings according to one or more criterion, generating one or more updated embeddings for each of the plurality of entities based on the refined initial embeddings of the plurality of entity attributes, and modifying the first set of embeddings based on the one or more updated embeddings.

Abstract: The disclosed method may include (1) determining a current physical state regarding an antenna assembly that includes (a) a sub-reflector that receives a wireless signal and reflects the wireless signal to a feed structure for processing, (b) a continuous antenna reflector that receives the wireless signal at a reflecting surface that reflects the wireless signal to the sub-reflector, where the current physical state is indicative of a current surface error over the reflecting surface relative to the sub-reflector, and (c) a backing structure coupled to a back surface of the continuous antenna reflector opposite the reflecting surface and having a plurality of actuators distributed over, and coupled to, the back surface, (2) operating each of the plurality actuators in a manner that reduces the current surface error based on the current physical state. Various other methods and systems are also disclosed.

Abstract: In one embodiment, a method for tensor data distribution using a direct-memory access agent includes generating, by a first controller, source addresses indicating locations in a source memory where portions of a source tensor are stored. A second controller may generate destination addresses indicating locations in a destination memory where portions of a destination tensor are to be stored. The direct-memory access agent receives a source address generated by the first controller and a destination address generated by the second controller and determines a burst size. The direct-memory access agent may issue a read request comprising the source address and the burst size to read tensor data from the source memory and may store the tensor data into an alignment buffer. The direct-memory access agent then issues a write request comprising the destination address and the burst size to write data from the alignment buffer into the destination memory.

Abstract: Systems and methods for digitally fetching music content may include (1) providing a user of a device with a music service accessible via the device, (2) receiving a digital user request to (i) select a music composition based on a digital image corresponding to one or more music compositions and (ii) digitally fetch the music composition for use via the music service, and, in response to receiving the digital user request, (3) analyzing the digital image and selecting the music composition based on the analysis of the digital image, and (4) digitally fetching the music composition for use via the music service. Various other methods, systems, and computer-readable media are also disclosed.

Abstract: Systems and methods for providing ephemeral streaming spaces may include (1) providing a user with a digital display of video cards, in which each video card within the digital display corresponds to a contact, of the user, designated as currently available for video conferencing, (2) receiving user input selecting a video card from the digital display, and (3) in response to receiving the user input, initiating a video conference between the user and a contact corresponding to the selected video card. Various other methods, systems, and computer-readable media are also disclosed.

Abstract: A system including a machine learning accelerator (MLA) hardware configured to perform machine-learning operations according to native instructions; an interpreter computing module configured to: generate, based on virtual instructions, machine language instructions configured to be processed by a processing hardware implementing the interpreter computing module; and cause the processing hardware to perform machine-learning operations according to the machine language instructions; and a compiler computing module associated with the MLA hardware, the compiler computing module configured to: receive instructions for performing an inference using a machine-learning model; based on the received instructions: generate the native instructions configured to be processed by the MLA hardware, the native instructions specifying first machine-learning operations associated with performing the inference; and generate the virtual instructions configured to be processed by the interpreter computing module, the virtual ins

Abstract: When a user creates a new video exchange session via an online system, the user specifies an activity for the video exchange session. The activity may specify a name for the video exchange session, as well as other thematic content (e.g., visual effects, backgrounds) displayed via the video exchange session. The online system trains a machine learned model that predicts a likelihood that other users will accept an invitation to join the video exchange session if a certain activity is specified for the video exchange session. The online system may then use these likelihoods to rank and sort the activities for the user or to highlight activities with threshold likelihoods.

Abstract: This disclosure covers systems and methods that both transmit a live video stream from a broadcaster device to viewer devices and relay video clips of reactions (by viewers to the live video stream) to the broadcaster device during presentation of the live video stream. In certain embodiments, the disclosed systems and methods facilitate viewers capturing video clips of reactions to a live video stream using a viewer device and, in turn, transmit video clips received from the viewer device to a broadcaster device during broadcast of the live video stream. For instance, in some embodiments, the systems and methods present the video clips of a viewer's reaction to the broadcaster device during a live-video-stream broadcast. Additionally, in some embodiments, the systems and methods present video clips of other viewers' reactions to a viewer device during a live-video-stream broadcast.

Abstract: In one embodiment, a method includes receiving, from a client system associated with a user, a user request comprising a reference to a target object, accessing visual data from the client system, wherein the visual data comprises images portraying the target object and one or more additional objects, and wherein attribute information of the target object is recorded in a multimodal dialog state, resolving the reference to the target object based on the attribute information recorded in the multimodal dialog state, determining relational information between the target object and one or more of the additional objects portrayed in the visual data, and sending, to the client system, instructions for presenting a response to the user request, wherein the response comprises the attribute information and the determined relational information.

Abstract: Systems, methods, and non-transitory computer readable media can provide one or more candidate stickers through a display screen of a computing device. A determination can be made that a user has selected a first sticker from the one or more candidate stickers. The first sticker can be inserted as a graphical overlay in a content item. The first sticker can reference first content.

Abstract: In one embodiment, a method includes, by a client system, receiving, at the client system, a first user input, processing by the client system, the first user input to provide an initial response by identifying one or more entities referenced by the first user input and providing, by the client system, the initial response, where the initial response includes a conversational filler referencing at least one of the one or more identified entities, processing the first user input to provide a complete response by identifying, by the client system, one or more intents and one or more slots associated with the first user input based on a semantic analysis by a natural-language understanding module, and providing, by the client system, the complete response subsequent to the initial response, where the complete response is based on the one or more intents and the one or more slots.