Google LLC (20240428056). PERFORMING TASKS USING GENERATIVE NEURAL NETWORKS

PERFORMING TASKS USING GENERATIVE NEURAL NETWORKS

Organization Name

Google LLC

Inventor(s)

Paul Kishan Rubenstein of Zurich (CH)

Matthew Sharifi of Kilchberg (CH)

Alexandru Tudor of Uitikon (CH)

Chulayuth Asawaroengchai of Zurich (CH)

Duc Dung Nguyen of Zurich (CH)

Marco Tagliasacchi of Ruvigliana (CH)

Neil Zeghidour of Paris (FR)

Zalán Borsos of Zurich (CH)

Christian Frank of Zurich (CH)

Dalia Salem Hassan Fahmy Elbadawy of Zurich (CH)

Hannah Raphaelle Muckenhirn of Zurich (CH)

Dirk Ryan Padfield of Seattle WA (US)

Damien Vincent of Zurich (CH)

Evgeny Kharitonov of Paris (FR)

Michelle Dana Tadmor of Tel Aviv (IL)

Mihajlo Velimirovic of Zurich (CH)

Feifan Chen of Seattle WA (US)

Victoria Zayats of Seattle WA (US)

PERFORMING TASKS USING GENERATIVE NEURAL NETWORKS

This abstract first appeared for US patent application 20240428056 titled 'PERFORMING TASKS USING GENERATIVE NEURAL NETWORKS

Original Abstract Submitted

methods, systems, and apparatus, including computer programs encoded on computer storage media, for performing tasks. one of the methods includes obtaining a sequence of input tokens, where each token is selected from a vocabulary of tokens that includes text tokens and audio tokens, and wherein the sequence of input tokens includes tokens that describe a task to be performed and data for performing the task; generating a sequence of embeddings by embedding each token in the sequence of input tokens in an embedding space; and processing the sequence of embeddings using a language model neural network to generate a sequence of output tokens for the task, where each token is selected from the vocabulary.