Abstract: Systems, methods, and non-transitory computer readable media configured to generate session information based on information regarding items of a plurality of item types associated with interactions performed by active users of a social networking system. A graph is generated based on the session information. At least a first item of the items is assigned to a cluster based on similarity between the item and the cluster. The cluster is provided to a recommender system to facilitate selection of relevant information for potential presentation to a user.

Abstract: In one embodiment, a method includes sending, by a client device, a request to a computer server system for a target structured document. In a first response phase, a first response portion from the computer server system is sent in response to the request. The first response portion includes a first resource associated with the target structured document, wherein a likelihood that the first resource will be included in a response to a future request for the target structured document exceeds a first predetermined threshold. Further in a second response phase, a second response portion from the computer server system is sent in response to the request. The second response portion includes a second resource associated with the target structured document, wherein a likelihood that the second resource will be included in a response to a future request for the target structured document exceeds a second predetermined threshold.

Abstract: A method and system for assembling a device by picking up semiconductor devices from a carrier substrate and placing the semiconductor devices onto a target substrate. The transfer of the semiconductor devices uses fluid as a transfer medium. The fluid enters a fluid channel of a pickup head, causing the pickup head to expand, make contact with, and attach to an aligned semiconductor device. After the semiconductor device is aligned with and placed onto the target substrate, at least a portion of the fluid is removed from the pickup head to release the semiconductor device onto the target substrate. The semiconductor device bonds to the target substrate.

Abstract: A circuit and method for calculating a non-linear function of floating-point numbers using hierarchical look-up tables are provided. The look-up tables are programmable to hold non-linear ranges of values for any of a variety of non-linear functions. The circuit includes computation modules in respective stages of a high-throughput computation pipeline. A first computation module in a first stage receives one or more floating-point numbers and, for each floating-point number, selects a first entry from a first look-up table based on the floating-point number. The first computation module then calculates and outputs a table index and a variable based on the first floating-point number and the first entry. The second compute module in a second stage, selects a second entry from a second look-up table based on the table index, and calculates and outputs an approximate value for the non-linear function using the variable and the second entry.

Abstract: An initiated chemical vapor deposition (i-CVD) method for forming a nanovoided polymeric material may include heating a mixture including a gaseous monomer, a gaseous polymerization initiator, and a solvent to form a polymeric thin film including the solvent and removing the solvent from the polymeric thin film to form a nanovoided thin film. Devices, including dielectric elastomer actuators, may be formed using the nanovoided polymeric material. Various other methods, systems, apparatuses, and materials are also disclosed.

Abstract: An optical element includes a nanovoided polymer layer having a first refractive index in an unactuated state and a second refractive index different than the first refractive index in an actuated state. Compression or expansion of the nanovoided polymer layer, for instance, can be used to reversibly control the size and shape of the nanovoids within the polymer layer and hence tune its refractive index over a range of values, e.g., during operation of the optical element. Various other apparatuses, systems, materials, and methods are also disclosed.

Abstract: Exemplary embodiments relate to improvements in spoken language understanding (SLU) systems. Conventionally, SLU systems include an automatic speech recognition (ASR) component configured to receive an input of audio data and to generate a textual representation of the audio data. Conventional SLU systems also include a natural language understanding (NLU) component configured to receive a text-based transcript and perform language-based tasks such as domain classification, intent determination, and slot-filling. However, these two components are typically trained separately based on different metrics. In real-world situations, errors in the ASR component propagate to the NLU component, which degrades the performance of the overall system. Exemplary embodiments described herein perform SLU in an end-to-end manner that infers semantic meaning directly from audio features without an intermediate text representation.

Abstract: An online system uses multiple machine learning models to select content for providing to a user of the online system. Specifically, the online system trains a general model that intakes a first set of features and outputs predictions at a general level. The online system further trains a residual model that intakes a second set of features. The residual model predicts a residual (e.g., an error) of the predictions outputted by the general model. Therefore, the predicted residual from the residual model is combined with the prediction from the general model in order to correct for the over-generality of the general model. The online system may use the combined prediction to send content to users.

Abstract: The disclosed computer-implemented method may include (1) maintaining a heterogeneous graph that represents (a) objects of a first type, (b) objects of a second type, and (c) relationships between the objects of the first type and objects of the second type, (2) using features of the objects of the first type to train a first embedding model to generate embeddings of the first type that (a) predict a label of objects of the first type and (b) predict, when combined with embeddings of the second type, graphical relationships in the graph, and (3) using features of each of the objects of the second type to train a second embedding model to generate the embeddings of the second type that predict, when combined with the embeddings of the first type, the graphical relationships in the graph. Various other methods, systems, and computer-readable media are also disclosed.

Abstract: An imaging waveguide for a visual display includes a substrate for guiding image light therein by total internal reflection (TIR). An input grating is supported by the substrate for coupling the image light into the imaging waveguide. An output grating is supported by the substrate and spaced apart from the input grating for coupling the image light guided in the substrate out of the imaging waveguide for observation by a user. A gap filling overcoat is formed on and within the output grating, but not on or within the input grating. The material is characterized by a refractive index between 1.40 and 1.80 at 500 nm, absorption between 0% and 1% in the visible region of the electromagnetic spectrum, and % haze between 0% and 0.2% in the visible region of the electromagnetic spectrum.

Abstract: An eye is illuminated with a first non-visible light wavelength and a second non-visible light wavelength. A first ocular image is captured from first reflected light having the first non-visible light wavelength and a second ocular image is captured from second reflected light having the second non-visible light wavelength.

Abstract: A client device accesses an online system using an authentication process when it is connected to a public network and not a private network. The client device requests access using an authentication broker via the public network. The authentication broker determines an authentication system and transmits identification information for the client device to the authentication system via the private network. The authentication broker receives an authentication vector generated by the authentication system via the private network, and receives a verification response from the client device via the public network. The verification response corresponds to a verification challenge generated based on the authentication vector by the authentication broker.

Abstract: In one embodiment, a computing system may access a first set of pixel values of an image having a bit depth. The first set of pixel values includes a first subset and a second subset. Each pixel value in the first subset is less than each pixel value in the second subset. The computing system may further determine an overflow range based on a pixel value overflow associated with the bit depth. The overflow range covers a second set of pixel values, which includes (1) the second subset and (2) an overflow representation of the first subset. In response to determining that the overflow range is less than a range covering the first set of pixel values, the computer system may then generate quantization levels corresponding to the overflow range and associate each pixel value of the second set of pixel values with one of the quantization levels.

Abstract: Systems, methods, and non-transitory computer readable media are configured to generate multiple channel embeddings for a page of a social networking system. The multiple channel embeddings can be mapped to a shared embedding space. A page embedding for the page of the social networking system can then be generated.

Abstract: Systems, methods, and non-transitory computer-readable media can provide a content feed comprising content items associated with a plurality of entities to be presented through a display screen of a computing device. A selection of an option to temporarily prevent content items satisfying at least one criterion from appearing in the content feed for a duration of time can be determined. At least one content item that satisfies the at least one criterion can be removed from the content feed.

Abstract: An optical device includes a spatial light modulator and an optical waveguide with a plurality of extraction features. The plurality of extraction features is positioned relative to the optical waveguide so that a respective extraction feature receives light, having propagated within the optical waveguide, in a first direction and directs a first portion of the light in a second direction distinct from the first direction to exit the optical waveguide and illuminate at least a portion of the spatial light modulator. The plurality of extraction features is also positioned relative to the optical waveguide so that a respective extraction feature directs a second portion, distinct from the first portion, of the light to undergo total internal reflection, thereby continuing to propagate within the optical waveguide.