Oriol Vinyals

This inventor holds 89 USPTO granted patents and 18 published patent applications, plus 2 CIPO patents and 19 EPO patents, primarily in Reinforcement Learning (CPC class G06N3-08). Top assignees: Google Inc., Deepmind Technologies Limited, Gdm Holding LLC. Active years: 2016-2026.

USPTO Granted Patents = 89 (Patent Applications = 18)

% Patents Active = 100.0

CIPO Patents = 2 (Patent Applications = 4)

EPO Patents = 19 (Patent Applications = 22)

Average Co-Inventor Count = 4.0

ph-index = 10

Forward Citations = 452(Granted Patents)

Forward Citations (Not Self Cited) = 340(Dec 10, 2025)

Inventors with similar research interests:

🔬 See more inventors

Location History:

Palo Alto, CA (US) (2017 - 2019)
Berkeley, CA (US) (2016 - 2020)
London, GB (2018 - 2024)

Company Filing History:

Deepmind Technologies Limited

Years Active: 2016-2026

Loading Chart...

Areas of Expertise:

Reinforcement Learning

Neural Networks

Multi-Agent Systems

Generative Models

Graph Representations

Sequence Generation

Audio Processing

Natural Language Processing

Machine Learning

Attention Mechanisms

Memory Adaptation

Transformers

89 patents (USPTO):Sort by Forward CitationsExplore Patents

1. 12718053 - Learning observation representations by predicting the future in latent space (2026-08-25),

2. 12700227 - Multimodal few-shot learning with frozen language models (2026-08-04),

3. 12688410 - Generating prediction outputs using dynamic graphs (2026-07-21),

4. 12670356 - Generating representations of input sequences using neural networks (2026-06-30),

5. 12632737 - Using hierarchical representations for neural network architecture searching (2026-05-19),

6. 12620408 - Generating audio using neural networks (2026-05-05),

7. 12572803 - Multi-agent reinforcement learning with matchmaking policies (2026-03-10),

8. 12547890 - Autoregressively generating sequences of data elements defining actions to be performed by an agent (2026-02-10),

9. 12536478 - Training distilled machine learning models (2026-01-27),

10. 12536408 - Universal transformers (2026-01-27),

11. 12535995 - Computer code generation from task descriptions using neural networks (2026-01-27),

12. 12518143 - Feedforward generative neural networks (2026-01-06),

13. 12518128 - Generative neural network systems for generating instruction sequences to control an agent performing a task (2026-01-06),

14. 12505346 - Autoregressively generating sequences of data elements defining actions to be performed by an agent (2025-12-23),

15. 12373695 - Auto-regressive neural network systems with a soft attention mechanism using support data patches (2025-07-29),

Title: Innovator Spotlight: Oriol Vinyals in the Realm of Machine Learning

Introduction: Oriol Vinyals is an esteemed inventor located in London, GB, with an impressive portfolio of 74 patents. His innovative contributions, particularly in the field of machine learning, have significantly advanced the capabilities of artificial intelligence.

Latest Patents: Among his recent groundbreaking patents is the method for machine learning systems that utilize memory-based parameter adaptation for fast and slower learning. This patent outlines a computer-implemented method of processing input data through a parametric model that includes both a first sub-model and a second sub-model. By generating a query data item from the input data with the first sub-model and retrieving relevant data point-value pairs from memory, the method optimizes the weights of the second sub-model, leading to efficient output data generation.

Another noteworthy patent by Vinyals is a neural network system designed for generating structured representations of objects through graph neural networks. This system enables the creation of a graph composed of nodes and edges, utilizing neural networks to represent probability distributions over sequences of decision-making in generating these nodes and edges. The ability to sample these distributions allows for intricate graph generation, opening new avenues in data representation and analysis.

Career Highlights: Oriol Vinyals has held significant positions at prominent companies, including Google Inc. and DeepMind Technologies Limited. His work at these organizations has placed him at the forefront of AI research and development, allowing him to contribute cutting-edge innovations to the field.

Collaborations: Throughout his career, Vinyals has collaborated with several notable experts, including Aaron Gerard Antonius Van Den Oord and Navdeep Jaitly. These partnerships have been instrumental in driving forward innovative projects and expanding the frontiers of machine learning.

Conclusion: With a remarkable number of patents and innovative contributions to machine learning, Oriol Vinyals stands as a significant figure in the technology landscape. His work not only influences current advancements but also sets the stage for future developments in artificial intelligence and its applications across various industries.