公开(公告)号：US20250103796A1

公开(公告)日：2025-03-27

申请号：US18846833

申请日：2023-02-23

Applicant: Google LLC

Inventor： Alvin Rajkomar ,  Eric Loreaux ,  Yuchen Liu ,  Ming-Jun Chen ,  Yi Zhang ,  Afroz Mohiuddin ,  Juraj Gottweis

IPC: G06F40/166 ,  G06V30/41 ,  G16H10/60 ,  G16H70/20

Abstract: Example embodiments relate to expanding textual content using transfer learning and iterative inference. An example method includes receiving, by a computing device, a snippet of text that contains one or more terms expressed using succinct representations. The method also includes performing an iterative expansion, by the computing device, using the snippet of text as an input snippet of text. The iterative expansion includes receiving, by the computing device, the input snippet of text. The iterative expansion also includes determining, by the computing device using a machine-learned model, a set of intermediate expanded snippets. Each of the intermediate expanded snippets has an associated score based on the machine-learned model. Additionally, the iterative expansion includes: (i) repeating, by the computing device, the iterative expansion using a first intermediate expanded snippet or a second intermediate expanded snippet as the input snippet or (ii) outputting the input snippet as a final expanded snippet.

公开(公告)号：US10592540B2

公开(公告)日：2020-03-17

申请号：US15195364

申请日：2016-06-28

Applicant: Google LLC

Inventor： Yi Liu ,  Preyas Popat ,  Nitin Gupta ,  Afroz Mohiuddin

IPC: G06F16/00 ,  G06F16/33 ,  G06F16/35 ,  G06F16/332

Abstract: Methods, systems, and apparatus, including computer programs encoded on computer storage media, for generating answers to answer-seeking queries. One of the methods includes receiving a query having multiple terms. The query is classified as an answer-seeking query of a particular question type, and one or more answer types associated with the particular question type are obtained. Search results satisfying the query are obtained, and a respective score is computed for each of one or more passages of text occurring in each document identified by the search results, wherein the score for each passage of text is based on how many of the one or more answer types match the passage of text. A presentation that includes information from one or more of the passages of text selected based on the respective score is provided in response to the query.

公开(公告)号：US20250087207A1

公开(公告)日：2025-03-13

申请号：US18736113

申请日：2024-06-06

Applicant: Google LLC

Inventor： Harshit Kharbanda ,  Jessica Lee ,  Christopher James Kelley ,  Fabian Roth ,  Dounia Berrada ,  Samer Hassan Hassan ,  Afroz Mohiuddin ,  Misha Khalman ,  Ali Essam Ali Elqursh ,  Belinda Luna Zeng

IPC: G10L15/183 ,  G06F16/583 ,  G06V10/778 ,  G06V30/14 ,  G06V30/148 ,  G10L15/22 ,  G10L15/30

Abstract: The present disclosure provides computer-implemented methods, systems, and devices for responding to requests associated with an image. A computing system obtains, wherein the image depicts a first set of textual content. The computing system determines one or more characteristics of the first set of textual content. The computing system determines a response type from a plurality of response types based on the one or more characteristics. The computing system generates a model input, wherein the model input comprises data descriptive of the first set of textual content and a prompt associated with the response type. The computing system provides providing the model input as an input to a machine-learned language model. The computing system receives a second set of text as an output of the machine-learned language model as a result of the machine-learned language model processing the model input. The computing system provides the second set of text for display to a user, wherein the second set of textual content is associated with the response type.

公开(公告)号：US12033620B1

公开(公告)日：2024-07-09

申请号：US18463951

申请日：2023-09-08

Applicant: Google LLC

Inventor： Harshit Kharbanda ,  Jessica Lee ,  Christopher James Kelley ,  Fabian Roth ,  Dounia Berrada ,  Samer Hassan Hassan ,  Afroz Mohiuddin ,  Mikhail Khalman ,  Ali Essam Ali Elqursh ,  Belinda Luna Zeng

IPC: G06F3/0483 ,  G06F16/30 ,  G06F16/33 ,  G06F16/583 ,  G06V10/778 ,  G06V30/14 ,  G06V30/148 ,  G10L15/183 ,  G10L15/22 ,  G10L15/30

CPC classification number: G10L15/183 ,  G06F16/5846 ,  G06V10/778 ,  G06V30/1456 ,  G06V30/153 ,  G10L15/22 ,  G10L15/30

Abstract: The present disclosure provides computer-implemented methods, systems, and devices for responding to requests associated with an image. A computing system obtains, wherein the image depicts a first set of textual content. The computing system determines one or more characteristics of the first set of textual content. The computing system determines a response type from a plurality of response types based on the one or more characteristics. The computing system generates a model input, wherein the model input comprises data descriptive of the first set of textual content and a prompt associated with the response type. The computing system provides providing the model input as an input to a machine-learned language model. The computing system receives a second set of text as an output of the machine-learned language model as a result of the machine-learned language model processing the model input. The computing system provides the second set of text for display to a user, wherein the second set of textual content is associated with the response type.

公开(公告)号：US20220253672A1

公开(公告)日：2022-08-11

申请号：US17666400

申请日：2022-02-07

Applicant: Google LLC

Inventor： Aakanksha Chowdhery ,  Afroz Mohiuddin ,  Henryk Michalewski ,  Jonni Miikka Kanerva ,  Lukasz Mieczyslaw Kaiser ,  Sebastian Dariusz Jaszczur ,  Wojciech Gajewski

IPC: G06N3/04

Abstract: Methods, systems, and apparatus, including computer programs encoded on a computer storage medium, for performing a machine learning task on a network input to generate a network output. In one aspect, one of the systems includes a neural network configured to perform the machine learning task, the neural network including one or more sparse attention layers.