Meta Platforms, Inc.

Meta Platforms Technologies, LLC

Facebook, Inc.

Honglei Liu

Wenhai Yang

David Scott Reiss

Kshitiz Malik

Nanshu Wang

Prince Gill

Pararth Paresh Shah

Xiaolei Li

Francislav Petrov Penov

Xiaohu Liu

Rajen Subba

Anuj Kumar

Seungwhan Moon

Eric Xiao

Lisa Xiaoyi Huang

Bing Liu

Hao Zhou

Yulong Qiu

Daniel Chai

Piyush Khemka

Jivko Dobrev

Hu Xu

Wenxuan Li

In one embodiment, a method includes training a target machine-learning model iteratively by accessing training data of content objects, training an intermediate machine-learning model that outputs contextual evaluation measurements based on the training data, generating state-indications associated with the training data, wherein the state-indications comprise user-intents, system actions, and user actions, training the target machine-learning model based on the contextual evaluation measurements, the state-indications, and an action set comprising possible system actions, extracting rules based on the target machine-learning model by a sequential pattern-mining model, generating synthetic training data based on the rules, updating the training data by adding the synthetic training data to the training data, determining if a completion condition is reached for the training, and if the completion condition is reached returning the target machine-learning model, else repeating the iterative training of the target machine-learning model.

Interpretability of deep reinforcement learning models in assistant systems

In one embodiment, a method includes receiving a user request from a client system associated with a user, generating a response to the user request which references one or more entities, generating a personalized recommendation based on the user request and the response, wherein the personalized recommendation references one or more of the entities of the response, and sending instructions for presenting the response and the personalized recommendation to the client system.

Personalized conversational recommendations by assistant systems

In one embodiment, a method includes maintaining a reminder store comprising one or more reminders for a first user, wherein each reminder is associated with an initial activation-condition, determining a context associated with the first user based on signals from a client system associated with the first user, determining a suppression-condition associated with at least one of the reminders is met based on the context, and suppressing the at least one reminder for the first user.

Suppressing reminders for assistant systems

In one embodiment, a method includes receiving a user request to create a reminder from a client system associated with a user, wherein the user request does not specify an activation-condition for the reminder, determining one or more proactive activation-conditions for the reminder, storing the reminder in a reminder store, receiving one or more inputs associated with the user, determining a user context associated with the user based on the one or more inputs, determining the one or more proactive activation-conditions for the reminder are satisfied based on the user context, and sending instructions for presenting the reminder to the user to the client system responsive to determining the one or more proactive activation-conditions are satisfied.

Generating proactive reminders for assistant systems

Exemplary embodiments relate to methods, mediums, and systems for moving language models from a server to the client device. Such embodiments may be deployed in an environment where the server is not able to provide modeling services to the clients, such as an end-to-end encrypted (E2EE) environment. Several different techniques are described to address issues of size and complexity reduction, model architecture optimization, model training, battery power reduction, and latency reduction.

Methods, mediums, and systems for providing a model for an end-user device

Methods, mediums, and systems for training a model

Methods, mediums, and systems for representing a model in a memory of device

Growing community of inventors

San Mateo, CA, United States of America

Honglei Liu