Abstract: Example embodiments of aspects of the present disclosure provide an example computer-implemented method for improved prompting of a machine-learned model. The example method can include obtaining an instructive sequence descriptive of an instructive query, an instructive response, and an instructive trace of intermediate states from the instructive query to the instructive response. The example method can include inputting, to a machine-learned model, the instructive sequence and an operative query, wherein the machine-learned model is configured to process the operative query with attention over the instructive sequence. The example method can include generating, using the machine-learned model and responsive to the operative query, an operative response.

Abstract: An example method for pretraining a machine-learned model is provided. The example method includes obtaining a plurality of different combinations of configuration parameters of a pretraining objective framework. The example method includes generating, using the pretraining objective framework, a plurality of corrupted training examples from one or more training examples, wherein the plurality of corrupted training examples are respectively generated according to the plurality of different combinations. The example method includes inputting the plurality of corrupted training examples into the machine-learned model, wherein the machine-learned model is configured to generate uncorrupted subportions corresponding to corrupted subportions of the corrupted training examples. The example method includes obtaining, from the machine-learned model, a plurality of outputs respectively generated by the machine-learned model based on the plurality of corrupted training examples.

Abstract: A system including a main neural network for performing one or more machine learning tasks on a network input to generate one or more network outputs. The main neural network includes a Mixture of Experts (MoE) subnetwork that includes a plurality of expert neural networks and a gating subsystem. The gating subsystem is configured to: apply a softmax function to a set of gating parameters having learned values to generate a respective softmax score for each of one or more of the plurality of expert neural networks; determine a respective weight for each of the one or more of the plurality of expert neural networks; select a proper subset of the plurality of expert neural networks; and combine the respective expert outputs generated by the one or more expert neural networks in the proper subset to generate one or more MoE outputs.

Abstract: Example aspects of the present disclosure are directed to systems and methods for performing automatic label smoothing of augmented training data. In particular, some example implementations of the present disclosure which in some instances can be referred to “AutoLabel” can automatically learn the labels for augmented data based on the distance between the clean distribution and augmented distribution. AutoLabel is built on label smoothing and is guided by the calibration-performance over a hold-out validation set. AutoLabel is a generic framework that can be easily applied to existing data augmentation methods, including AugMix, mixup, and adversarial training, among others. AutoLabel can further improve clean accuracy, as well as the accuracy and calibration over corrupted datasets. Additionally, AutoLabel can help adversarial training by bridging the gap between clean accuracy and adversarial robustness.

Abstract: Implementations provide an improved system for presenting search results based on entity associations of the search items. An example method includes generating first-level clusters of items responsive to a query, each cluster representing an entity in a knowledge base and including items mapped to the entity, merging the first-level clusters based on entity ontology relationships, applying hierarchical clustering to the merged clusters, producing final clusters, and initiating display of the items according to the final clusters. Another example method includes generating first-level clusters from items responsive to a query, each cluster representing an entity in a knowledge base and including items mapped to the entity, producing final clusters by merging the first-level clusters based on an entity ontology and an embedding space that is generated from an embedding model that uses the mapping, and initiating display of the items responsive to the query according to the final clusters.

Abstract: Implementations provide an improved system for presenting search results based on entity associations of the search items. An example method includes generating first-level clusters of items responsive to a query, each cluster representing an entity in a knowledge base and including items mapped to the entity, merging the first-level clusters based on entity ontology relationships, applying hierarchical clustering to the merged clusters, producing final clusters, and initiating display of the items according to the final clusters. Another example method includes generating first-level clusters from items responsive to a query, each cluster representing an entity in a knowledge base and including items mapped to the entity, producing final clusters by merging the first-level clusters based on an entity ontology and an embedding space that is generated from an embedding model that uses the mapping, and initiating display of the items responsive to the query according to the final clusters.

Abstract: Implementations provide an improved system for presenting search results based on entity associations of the search items. An example method includes, for each of a plurality of crowdsource workers, initiating display of a first randomly selected cluster set from a plurality of cluster sets to the crowdsource worker. Each cluster set represents a different clustering algorithm applied to a set of search items responsive to a query. The method also includes receiving cluster ratings for the first cluster set from the crowdsource worker and calculating a cluster set score for the first cluster set based on the cluster ratings. This is repeated for remaining cluster sets in the plurality of cluster sets. The method also includes storing a cluster set definition for a highest scoring cluster set, associating the cluster set definition with the query, and using the definition to display search items responsive to the query.

Abstract: Described is a technique for managing a workflow of human intelligence tasks based on task performance. When a large batch of tasks is performed continuously by a worker, task performance may decline. To lessen these consequences and improve overall task performance, the techniques described herein may adjust the type of tasks provided during a workflow. These adjustments may include providing a workflow interruption in the form of a different type of task or a break activity. These interruptions may switch between conceptual and perceptual activities in order to refresh the user and aid in alleviating the negative consequences of repetitive tasks such as physical and cognitive fatigue.

Abstract: A system for creating an activity-based social includes receiving information from a computing device of a participant in an activity, and determining that the information qualifies the participant for membership in a social network associated with the activity. The system also includes associating the participant with the social network, and enabling access, by the participant, to an electronic portal that provides access to the social network.

Abstract: Systems and methods for determining a group recommendation of an object, such as a restaurant, movie, or other object, from a plurality of candidate objects based on user comparisons of characteristic traits of the candidate objects are provided. In particular, keywords associated with characteristic traits are identified. The keywords are then presented to members of the group as a series of selection queries. The selection queries require a user to select or rank the keywords based on user preferences. The responses to the selection queries are used to generate a ranking score for each of the plurality of candidate objects and to select one or more of the candidate objects to recommend to the group.