We introduce Jukebox, a model that generates music with singing in the raw audio domain. We tackle the long context of raw audio using a multi-scale VQ-VAE to compress it to discrete codes, and modeling those using autoregressive Transformers. We show that the combined model at scale can generate high-fidelity and diverse songs with coherence up to multiple minutes. We can condition on artist and genre to steer the musical and vocal style, and on unaligned lyrics to make the singing more controllable. We are releasing thousands of non cherry-picked samples at https://jukebox.openai.com, along with model weights and code at https://github.com/openai/jukebox

Our effort is toward unifying GAN and DRL algorithms into a unifying AI model (AGI or general-purpose AI or artificial general intelligence which has general-purpose applications to: (A) offline learning (of stored data) like GAN in (un/semi-/fully-)SL setting such as big data analytics (mining) and visualization; (B) online learning (of real or simulated devices) like DRL in RL setting (with/out environment reward) such as (real or simulated) robotics and control; Our core proposal is adding an (generative/predictive) environment model to the actor-critic (model-free) architecture which results in a model-based actor-critic architecture with temporal-differencing (TD) error and an episodic memory. The proposed AI model is similar to (model-free) DDPG and therefore it's called model-based DDPG. To evaluate it, we compare it with (model-free) DDPG by applying them both to a variety (wide range) of independent simulated robotic and control task environments in OpenAI Gym and Unity Agents. Our initial limited experiments show that DRL and GAN in model-based actor-critic results in an incremental goal-driven intellignce required to solve each task with similar performance to (model-free) DDPG. Our future focus is to investigate the proposed AI model potential to: (A) unify DRL field inside AI by producing competitive performance compared to the best of model-based (PlaNet) and model-free (D4PG) approaches; (B) bridge the gap between AI and robotics communities by solving the important problem of reward engineering with learning the reward function by demonstration;

Recent work in adversarial attacks has developed provably robust methods for training deep neural network classifiers. However, although they are often mentioned in the context of robustness, deep generative models themselves have received relatively little attention in terms of formally analyzing their robustness properties. In this paper, we propose a method for training provably robust generative models, specifically a provably robust version of the variational auto-encoder (VAE). To do so, we first formally define a (certifiably) robust lower bound on the variational lower bound of the likelihood, and then show how this bound can be optimized during training to produce a robust VAE. We evaluate the method on simple examples, and show that it is able to produce generative models that are substantially more robust to adversarial attacks (i.e., an adversary trying to perturb inputs so as to drastically lower their likelihood under the model).

OpenAI has recently launched Microscope in order to help researchers understand the architecture and behaviour of neural networks in a better way. According to the company, Microscope is a library of neuron visualisations starting with nine popular or heavily neural networks -- a vast collection encompasses millions of images. As the name suggests and similar to its usage, in a laboratory, Microscope has been designed to help AI researchers better understand the complex structure of neural networks with tens of thousands of neurons. In the OpenAI Microscope website, it has been stated that the "OpenAI Microscope is a collection of visualisations of every significant layer and neuron of several common "model organisms" which are often studied in interpretability. Microscope makes it easier to analyse the features that form inside these neural networks, and we hope it will help the research community as we move towards understanding these complicated systems."

This week, we witnessed open-source tools focusing mostly on making models lighter and explainable. OpenAI, especially, has come up with an interesting tool to promote the interpretability of ML models. Furthermore, TensorFlow has made it even more simple for developers to execute their models. Let us take a look at top AI news for developers from this week. OpenAI Microscope tool is a collection of visualisations of every significant layer and neuron of eight vision'model organisms', which are often studied in interpretability.

OpenAI yesterday unveiled its Open AI Microscope, which provides visualizations of every significant layer and neuron in eight of today's most popular computer vision (CV) models. Interactions between neurons indicate the abilities of neural networks, and with machine learning trending toward increasingly complicated neural networks it is important for researchers to be able to quickly and easily conduct a closer inspection of these thousands of interactions. This is where AI Microscope comes in. Just as biologists gain insights into organisms by putting model specimens under their microscopes, AI Microscope was designed to help researchers analyze the features that form inside leading CV models. OpenAI explains that its Microscope models are composed of a graph of nodes -- neural network layers connected via edges.

OpenAI today launched Microscope, a library of neuron visualizations starting with nine popular or heavily neural networks. In all, the collection encompasses millions of images. Like a microscope can do in a laboratory, Microscope is made to help AI researchers better understand the architecture and behavior of neural networks with tens of thousands of neurons. Initial models in Microscope include historically important and commonly studied computer vision models like AlexNet, 2012 winner of the now retired ImageNet challenge. AlexNet has been cited over 50,000 times in research.

Introducing OpenAI Microscope: a collection of visualizations of every layer and neuron in eight vision "model organisms" often studied in interpretability.

Insilico Medicine is pleased to announce that it has entered into a research collaboration with Boehringer Ingelheim to utilize Insilico's generative machine learning technology and proprietary Pandomics Discovery Platform with the aim of identifying potential therapeutic targets implicated in a variety of diseases. "Insilico Medicine is very impressed with the Research Beyond Borders group at Boehringer Ingelheim capabilities in the search of potential drug targets. In this collaboration, Insilico will provide additional AI capabilities to discover novel targets for a variety of diseases to benefit the patients worldwide. We are very happy to partner with such an advanced group," said Alex Zhavoronkov, PhD, founder, and CEO of Insilico Medicine. "We believe that Insilico's exclusive Pandomics platform will provide huge boost to our ability to explore and identify drug targets. We look forward to using AI to significantly improve the drug discovery process and contribute to human health," said from Dr. Weiyi Zhang, Head of External Innovation Hub, Boehringer Ingelheim Greater China.

Easy tips to train your Reinforcement Learning AI with Unity3D using the ML-Agents Framework. My name is Sebastian Schuchmann, AI enthusiast from Germany and we are going to cover simple, beginner-friendly ways to improve your Machine Learning process. The Algorithm used is called PPO and was developed by OpenAI (founded by Elon Musk). After watching this video you will hopefully be able to train an Artificial Intelligence to crack your favorite game. I am very curious about what you guys will create!