Abstract: There are provided solutions for multiagent debate. In a method, a first and a second response representation are determined by a first and a second agent in a plurality of agents based on a first query representation for a query in a natural language, respectively, and the first and second response representations are convertible to a first and a second answer to the query in the natural language, respectively. A second query representation is obtained based on the first query representation, and at least one of the first and second response representations. A response representation is determined based on the second query representation by at least one of the first and second agents, and the response representation is convertible to an answer to the query in the natural language. These agents communicate in an embedding space without a conversion between a natural language format and an embedding format.

Abstract: A method performed by at least one processing device in an illustrative embodiment comprises applying a first image and a message to an encoder of a steganographic encoder-decoder neural network, generating in the encoder, based at least in part on the first image and the message, a perturbed image containing the message, decoding the perturbed image in a decoder of the steganographic encoder-decoder neural network, and providing information characterizing the decoded perturbed image to the encoder. The generating, decoding and providing are iteratively repeated, with different perturbations being determined in the encoder as a function of respective different instances of the provided information, until the decoded perturbed image meets one or more specified criteria relative to the message. The perturbed image corresponding to the decoded perturbed image that meets the one or more specified criteria relative to the message is output as a steganographic image containing the message.

Abstract: An electronic device includes a housing defining an internal volume, a front opening, and a rear opening. The electronic device can include a display component disposed at the front opening and a rear cover disposed at the rear opening. A logic board can be disposed in the internal volume. The device can also include a thin film thermopile including a cold junction bonded to the logic board and a hot junction bonded to the rear cover.

Abstract: Embodiments of the disclosed technologies receive a first-party trained model and a first-party data set from a first-party system into a protected environment, receive a first third-party data set into the protected environment, and, in a data clean room, joining the first-party data set and the first third-party data set to create a joint data set for the particular segment, tuning a first-party trained model with the joint data set to create a third-party tuned model, sending model parameter data learned in the data clean room as a result of the tuning to an aggregator node, receiving a globally tuned version of the first-party trained model from the aggregator node, applying the globally tuned version of the first-party trained model to a second third-party data set to produce a scored third-party data set, and providing the scored third-party data set to a content distribution service of the first-party system.

Abstract: The disclosed embodiments provide a system for streamlining machine learning. During operation, the system determines a resource overhead for a baseline version of a machine learning model that uses a set of features to produce entity rankings and a number of features to be removed to lower the resource overhead to a target resource overhead. Next, the system calculates importance scores for the features, wherein each importance score represents an impact of a corresponding feature on the entity rankings. The system then identifies a first subset of the features to be removed as the number of features with lowest importance scores and trains a simplified version of the machine learning model using a second subset of the features that excludes the first subset of the features. Finally, the system executes the simplified version to produce new entity rankings.

Abstract: The disclosed embodiments provide a system for processing data. During operation, the system obtains a feature configuration for a feature. Next, the system obtains, from the feature configuration, an anchor containing metadata for accessing the feature in an environment. The system then uses one or more attributes of the anchor to retrieve one or more feature values of the feature from the environment. Finally, the system provides the one or more feature values for use with one or more machine-learning models.

Abstract: Technologies for monitoring performance of a machine learning model include receiving, by an unsupervised anomaly detection function, digital time series data for a feature metric; where the feature metric is computed for a feature that is extracted from an online system over a time interval; where the machine learning model is to produce model output that relates to one or more users' use of the online system; using the unsupervised anomaly detection function, detecting anomalies in the digital time series data; labeling a subset of the detected anomalies in response to a deviation of a time-series prediction model from a predicted baseline model exceeding a predicted deviation criterion; creating digital output that identifies the feature as associated with the labeled subset of the detected anomalies; causing, in response to the digital output, a modification of the machine learning model.

Abstract: The disclosed embodiments provide a system for streamlining machine learning. During operation, the system determines a resource overhead for a baseline version of a machine learning model that uses a set of features to produce entity rankings and a number of features to be removed to lower the resource overhead to a target resource overhead. Next, the system calculates importance scores for the features, wherein each importance score represents an impact of a corresponding feature on the entity rankings. The system then identifies a first subset of the features to be removed as the number of features with lowest importance scores and trains a simplified version of the machine learning model using a second subset of the features that excludes the first subset of the features. Finally, the system executes the simplified version to produce new entity rankings.

Abstract: In an example embodiment, a GLMix model is utilized that models viewers and actors of feed items. This allows for random effects of individual viewers and actors to be taken into account without introducing biases. Additionally, in an example embodiment, predictions/recommendations are made more accurate by using three models, which are then combined, instead of a single GLMix model. Each of these models has different granularities and dimensions. A global model may model the similarity between user attributes (e.g., from the member profile or activity history) and item attributes. A per-viewer model may model user attributes and activity history of actors on feed items. A per-actor model may model user attributes and activity history of the viewers of feed items. The per-actor model may therefore, rely on information regarding how and what type of viewers interacted with items acted on by the particular actor.

Abstract: A news feed system provided with an on-line social network system determines that a news feed is to be constructed for a viewer. The news feed system accesses the viewer's profile and other information associated with the viewer, accesses an inventory of activities that have been identified as potentially of interest to the viewer, and calculates relevance score for each item inventory of activities using the combined predictions methodology. The activities are then arranged for presentation to the viewer via a news feed web page, using respective calculated relevance scores.

Abstract: Technologies for monitoring performance of a machine learning model include receiving, by an unsupervised anomaly detection function, digital time series data for a feature metric; where the feature metric is computed for a feature that is extracted from an online system over a time interval; where the machine learning model is to produce model output that relates to one or more users' use of the online system; using the unsupervised anomaly detection function, detecting anomalies in the digital time series data; labeling a subset of the detected anomalies in response to a deviation of a time-series prediction model from a predicted baseline model exceeding a predicted deviation criterion; creating digital output that identifies the feature as associated with the labeled subset of the detected anomalies; causing, in response to the digital output, a modification of the machine learning model.

Abstract: Techniques for improving the accuracy, scalability, and efficiency of machine-learning models for selecting digital content items for display within a graphical user interface are disclosed herein.

Abstract: The disclosed embodiments provide a system for processing data. During operation, the system obtains a feature configuration for a feature. Next, the system obtains, from the feature configuration, an anchor containing metadata for accessing the feature in an environment. The system then uses one or more attributes of the anchor to retrieve one or more feature values of the feature from the environment. Finally, the system provides the one or more feature values for use with one or more machine-learning models.

Abstract: The disclosed embodiments provide a system for processing data. During operation, the system obtains feature configurations for a set of features. Next, the system obtains, from the feature configurations, an anchor containing metadata for accessing a first feature in an environment and a feature derivation for generating a second feature from the first feature. The system then uses the anchor to retrieve feature values of the first feature from the environment and uses the feature derivation to generate additional feature values of the second feature from the feature values of the first feature. Finally, the system provides the additional feature values for use with one or more machine learning models.

Abstract: In an example embodiment, a GLMix model is utilized that models viewers and actors of feed items. This allows for random effects of individual viewers and actors to be taken into account without introducing biases. Additionally, in an example embodiment, predictions/recommendations are made more accurate by using three models, which are then combined, instead of a single GLMix model. Each of these models has different granularities and dimensions. A global model may model the similarity between user attributes (e.g., from the member profile or activity history) and item attributes. A per-viewer model may model user attributes and activity history of actors on feed items. A per-actor model may model user attributes and activity history of the viewers of feed items. The per-actor model may therefore, rely on information regarding how and what type of viewers interacted with items acted on by the particular actor.

Abstract: A system and method for personalizing cross session diversity is disclosed. The system receives a member opportunity request. In response, the system generates a list of members in response to the received member opportunity request, wherein the list of members is determined based on member profile data stored at a social networking system. For each member in the generated list of members, the system generates a profile value score based on the stored member profile data. The system ranks the members of the generated list at least in part based on the generated profile value scores. The system then selects one or more members in the list of members based on the ranking of members in the generated list.

Abstract: A news feed system provided with an on-line social network system determines that a news feed is to be constructed for a viewer. The news feed system accesses the viewer's profile and other information associated with the viewer, accesses an inventory of activities that have been identified as potentially of interest to the viewer, and calculates relevance score for each item inventory of activities using the combined predictions methodology. The activities are then arranged for presentation to the viewer via a news feed web page, using respective calculated relevance scores.

Abstract: A system and method for personalizing cross session diversity is disclosed. The system receives a member opportunity request. In response, the system generates a list of members in response to the received member opportunity request, wherein the list of members is determined based on member profile data stored at a social networking system. For each member in the generated list of members, the system generates a profile value score based on the stored member profile data. The system ranks the members of the generated list at least in part based on the generated profile value scores. The system then selects one or more members in the list of members based on the ranking of members in the generated list.