San Francisco, CA, United States of America

James Bradbury

This inventor holds 13 USPTO granted patents and 1 published patent application, plus 4 CIPO patents and 1 EPO patent. Top assignees: Salesforce Com, Inc., Salesforce, Inc.. Active years: 2020-2025.

USPTO Granted Patents = 13 (Patent Applications = 1)

% Patents Active = 100.0

CIPO Patents = 4 (Patent Applications = 2)

EPO Patents = 1 (Patent Applications = 3)

Average Co-Inventor Count = 2.9

ph-index = 4

Forward Citations = 95(Granted Patents)

Location History:

Mountain View, CA (US) (2020 - 2023)
San Francisco, CA (US) (2020 - 2024)

Company Filing History:

Years Active: 2020-2025

Loading Chart...

13 patents (USPTO):Sort by Forward CitationsExplore Patents

1. 12235790 - Deep neural network-based decision network (2025-02-25),

2. 12198047 - Quasi-recurrent neural network based encoder-decoder model (2025-01-14),

3. 12014257 - Domain specific language for generation of recurrent neural network architectures (2024-06-18),

4. 11822897 - Systems and methods for structured text translation with tag alignment (2023-11-21),

5. 11580359 - Pointer sentinel mixture architecture (2023-02-14),

6. 11537801 - Structured text translation (2022-12-27),

7. 11520998 - Neural machine translation with latent tree attention (2022-12-06),

8. 11354565 - Probability-based guider (2022-06-07),

9. 11250311 - Deep neural network-based decision network (2022-02-15),

10. 11080595 - Quasi-recurrent neural network based encoder-decoder model (2021-08-03),

11. 10963652 - Structured text translation (2021-03-30),

12. 10565493 - Pointer sentinel mixture architecture (2020-02-18),

13. 10565318 - Neural machine translation with latent tree attention (2020-02-18),

Title: Innovations of James Bradbury

Introduction

James Bradbury is a prominent inventor based in San Francisco, CA. He has made significant contributions to the field of artificial intelligence and natural language processing. With a total of 13 patents, his work continues to influence the technology landscape.

Latest Patents

One of his latest patents is a deep neural network-based decision network. This technology proposes a combination of a computationally cheap, less-accurate bag of words (BoW) model and a computationally expensive, more-accurate long short-term memory (LSTM) model for natural processing tasks such as sentiment analysis. The BoW model is referred to as 'skimming', while the LSTM model is termed “reading”. The technology also introduces a probability-based guider (PBG) that uses a probability thresholding strategy to determine whether to invoke the LSTM model for reliably classifying a sentence as positive or negative. Another notable patent is the quasi-recurrent neural network-based encoder-decoder model. This model alternates convolutional layers that apply in parallel across timesteps and minimalist recurrent pooling layers that apply in parallel across feature dimensions.

Career Highlights

James has worked with notable companies such as Salesforce, Inc. His experience in these organizations has allowed him to refine his skills and contribute to groundbreaking innovations in the tech industry.

Collaborations

Throughout his career, James has collaborated with esteemed colleagues, including Richard Socher and Caiming Xiong. These partnerships have further enhanced his work and led to significant advancements in artificial intelligence.

Conclusion

James Bradbury's contributions to technology, particularly in artificial intelligence and natural language processing, are noteworthy. His innovative patents and collaborations reflect his commitment to advancing the field.