Salesforce Com, Inc.

Lichao Sun

Caiming Xiong

Jia Li

Richard Socher

Yingbo Zhou

Kazuma Hashimoto

A method for evaluating robustness of one or more target neural network models using natural typos. The method includes receiving one or more natural typo generation rules associated with a first task associated with a first input document type, receiving a first target neural network model, and receiving a first document and corresponding its ground truth labels. The method further includes generating one or more natural typos for the first document based on the one or more natural typo generation rules, and providing, to the first target neural network model, a test document generated based on the first document and the one or more natural typos as an input document to generate a first output. A robustness evaluation result of the first target neural network model is generated based on a comparison between the output and the ground truth labels.

Robustness evaluation via natural typos

Systems and methods are provided for near-zero-cost (NZC) query framework or approach for differentially private deep learning. To protect the privacy of training data during learning, the near-zero-cost query framework transfers knowledge from an ensemble of teacher models trained on partitions of the data to a student model. Privacy guarantees may be understood intuitively and expressed rigorously in terms of differential privacy. Other features are also provided.

Near-zero-cost differentially private deep learning with teacher ensembles

A method for training parameters of a student model includes receiving one or more teacher models trained using sensitive data. Each teacher model includes one or more intermediate layers and a prediction layer coupled to the one or more intermediate layers. The method includes receiving, from the one or more teacher models, one or more intermediate layer outputs and one or more prediction layer outputs respectively based on public data. Student model training is performed to train parameters of the student model based on the intermediate layer outputs and prediction layer outputs of the one or more teacher models.

Private deep learning

Approaches for private and interpretable machine learning systems include a system for processing a query. The system includes one or more teacher modules for receiving a query and generating a respective output, one or more privacy sanitization modules for privacy sanitizing the respective output of each of the one or more teacher modules, and a student module for receiving a query and the privacy sanitized respective output of each of the one or more teacher modules and generating a result. Each of the one or more teacher modules is trained using a respective private data set. The student module is trained using a public data set. In some embodiments, human understandable interpretations of an output from the student module is provided to a model user.

Growing community of inventors

Chicago, IL, United States of America

Lichao Sun