Interpreting Deep Neural Networks using Cognitive Psychology
Deep neural networks have learnt to do an amazing array of tasks - from recognising and reasoning about objects in images to playing Atari and Go at super-human levels. As these tasks and network architectures become more complex, the solutions that neural networks learn become more difficult to understand.This is known as the ‘black-box’ problem, and it is becoming increasingly important as neural networks are used in more and more real world applications. At DeepMind, we are working to expand the toolkit for understanding and interpreting these systems. One potentially valuable tool is cognitive psychology.
作者：DeepMind 来源：DeepMind
发布时间：2017-06-28 01:16:00
Enhancing patient safety at Taunton and Somerset NHS Foundation Trust
We’re delighted to announce our first partnership outside of London to help doctors and nurses break new ground in the NHS’s use of digital technology. Streams is our secure mobile app that helps doctors and nurses give faster urgent care to patients showing signs of deterioration by giving them the right information more quickly. Over the next five years, we’ll be rolling it out at Taunton and Somerset NHS Foundation Trust as part of a new partnership.
作者：DeepMind 来源：DeepMind
发布时间：2017-06-21 17:30:00
Learning through human feedback
A central question in technical AI safety is how to tell an algorithm what we want it to do. Working with OpenAI, we demonstrate a novel system that allows a human with no technical experience to teach an AI how to perform a complex task, such as manipulating a simulated robotic arm.
作者：DeepMind 来源：DeepMind
发布时间：2017-06-12 23:00:00
A neural approach to relational reasoning
A key challenge in developing artificial intelligence systems with the flexibility and efficiency of human cognition is giving them an ability for relational reasoning - drawing logical conclusions about how physical objects, sentences, or even abstract ideas are related to one another. In two new papers, we explore the ability for deep neural networks to perform complicated relational reasoning with unstructured data. In A Simple Neural Network Module for Relational Reasoning we describe a Relation Network and show that it can perform at superhuman levels on a challenging task. While in Visual Interaction Networks we describe a general purpose model that can predict the future state of a physical object based purely on visual observations.
作者：DeepMind 来源：DeepMind
发布时间：2017-06-06 15:59:00
AlphaGo's next move
This week’s series of thrilling games with the world’s best players, in the country where Go originated, has been the highest possible pinnacle for AlphaGo as a competitive program. For that reason, the Future of Go Summit is our final match event with AlphaGo. To mark the end of the event we wanted to give a special gift to fans of Go around the world, publishing a special set of 50 AlphaGo vs AlphaGo games, played at full length time controls, which we believe contain many new and interesting ideas and strategies for the Go community to explore.
作者：DeepMind 来源：DeepMind
发布时间：2017-05-27 16:00:00
Exploring the mysteries of Go with AlphaGo and China's top players
We’re collaborating with the China Go Association and Chinese Government to bring AlphaGo, China’s top Go players, and leading AI experts together for the “Future of Go Summit.”
作者：DeepMind 来源：DeepMind
发布时间：2017-04-10 16:01:00
Innovations of AlphaGo
One of the great promises of AI is its potential to help us unearth new knowledge in complex domains. We’ve already seen exciting glimpses of this, when our algorithms found ways to dramatically improve energy use in data centres - as well as of course with our program AlphaGo. Since its historic success in Seoul last March, AlphaGo has heralded a new era for the ancient game of Go. Thanks to AlphaGo's creative and intriguing revelations, players of all levels have been inspired to test out new moves and strategies of their own, often re-evaluating centuries of inherited knowledge in the process. Ahead of ‘The Future of Go Summit in Wuzhen’, we summarise some recent examples of AlphaGo’s strategic and tactical innovations, and the new insights they have revealed.
作者：DeepMind 来源：DeepMind
发布时间：2017-04-10 16:00:00
Open sourcing Sonnet - a new library for constructing neural networks
We have found that the flexibility and adaptiveness of TensorFlow lends itself to building higher level frameworks for specific purposes, and we’ve written one for quickly building neural network modules with TF. We are actively developing this codebase, but what we have so far fits our research needs well, and we’re excited to announce that today we are open sourcing it. We call this framework Sonnet.
作者：DeepMind 来源：DeepMind
发布时间：2017-04-07 20:00:00
Distill: Communicating the science of machine learning
Today sees the launch of Distill - a new, independent, web-based medium for clear and open - demystified - machine learning research, comprising a journal, prizes and tools to create interactive essays. DeepMind is a proud contributing sponsor of the Distill prize, an annual prize aimed at recognising outstanding work communicating and refining ideas in machine learning, and Shakir Mohamed is a member of the journal's steering committee.
作者：DeepMind 来源：DeepMind
发布时间：2017-03-21 01:00:00
Enabling Continual Learning in Neural Networks
Computer programs that learn to perform tasks also typically forget them very quickly. In our latest paper, published in PNAS, we show that the learning rule can be modified so that a program can remember old tasks when learning a new one. In time, this approach could help us build problem-solving systems that can learn more flexibly and efficiently.
作者：DeepMind 来源：DeepMind
发布时间：2017-03-13 21:16:00
Trust, confidence and Verifiable Data Audit
Imagine a service that could give mathematical assurance about what is happening with each individual piece of personal data, without possibility of falsification or omission. Imagine the ability for the inner workings of that system to be checked in real-time, to ensure that data is only being used as it should be. Imagine that the infrastructure powering this was freely available as open source, so any organisation in the world could implement their own version if they wanted to. The working title for this project is “Verifiable Data Audit”, and we’re really excited to share more details about what we’re planning to build!
作者：DeepMind 来源：DeepMind
发布时间：2017-03-09 19:00:00
A milestone for DeepMind Health and Streams
In November we announced a groundbreaking five year partnership with the Royal Free London to deploy and expand on Streams, our secure clinical app that aims to improve care by getting the right information to the right clinician at the right time. The first version of Streams has now been deployed at the Royal Free and we’re delighted that the early feedback from nurses, doctors and patients has so far been really positive. Some of the nurses using Streams at the hospital estimate that the app is saving them up to two hours per day, giving them more time to spend with patients in need.
作者：DeepMind 来源：DeepMind
发布时间：2017-02-27 18:50:00
Understanding Agent Cooperation
We employ deep multi-agent reinforcement learning to model the emergence of cooperation. The new notion of sequential social dilemmas allows us to model how rational agents interact, and arrive at more or less cooperative behaviours depending on the nature of the environment and the agents’ cognitive capacity.
作者：DeepMind 来源：DeepMind
发布时间：2017-02-09 18:00:00
Our collaborations with academia to advance the field of AI
Our co-founder and CEO, Demis Hassabis, explains DeepMind's collaborations with academia - including a new machine learning training module at UCL - and why sharing talent, expertise and breakthroughs can help us all make better progress in the development of artificial intelligence and its application for positive social benefit.
作者：DeepMind 来源：DeepMind
发布时间：2017-01-23 18:00:00
DeepMind’s work in 2016: a round-up
Our founders take a look back at an exciting year, including the publication of two Nature papers, the success of AlphaGo, and the first signs of positive real-world impact in our work with Google's data centre team.
作者：DeepMind 来源：DeepMind
发布时间：2017-01-03 23:00:00
