Bayesian inverse reinforcement learning (IRL) methods are ideal for safe imitation learning, as they allow a learning agent to reason about reward uncertainty and the safety of a learned policy. However, Bayesian IRL is computationally intractable for high-dimensional problems because each sample from the posterior requires solving an entire Markov Decision Process (MDP). While there exist non-Bayesian deep IRL methods, these methods typically infer point estimates of reward functions, precluding rigorous safety and uncertainty analysis. We propose Bayesian Reward Extrapolation (B-REX), a highly efficient, preference-based Bayesian reward learning algorithm that scales to high-dimensional, visual control tasks. Our approach uses successor feature representations and preferences over demonstrations to efficiently generate samples from the posterior distribution over the demonstrator's reward function without requiring an MDP solver. Using samples from the posterior, we demonstrate how to calculate high-confidence bounds on policy performance in the imitation learning setting, in which the ground-truth reward function is unknown. We evaluate our proposed approach on the task of learning to play Atari games via imitation learning from pixel inputs, with no access to the game score. We demonstrate that B-REX learns imitation policies that are competitive with a state-of-the-art deep imitation learning method that only learns a point estimate of the reward function. Furthermore, we demonstrate that samples from the posterior generated via B-REX can be used to compute high-confidence performance bounds for a variety of evaluation policies. We show that high-confidence performance bounds are useful for accurately ranking different evaluation policies when the reward function is unknown. We also demonstrate that high-confidence performance bounds may be useful for detecting reward hacking.

In principle, meta-reinforcement learning algorithms leverage experience across many tasks to learn fast reinforcement learning (RL) strategies that transfer to similar tasks. However, current meta-RL approaches rely on manually-defined distributions of training tasks, and hand-crafting these task distributions can be challenging and time-consuming. Can "useful" pre-training tasks be discovered in an unsupervised manner? We develop an unsupervised algorithm for inducing an adaptive meta-training task distribution, i.e. an automatic curriculum, by modeling unsupervised interaction in a visual environment. The task distribution is scaffolded by a parametric density model of the meta-learner's trajectory distribution. We formulate unsupervised meta-RL as information maximization between a latent task variable and the meta-learner's data distribution, and describe a practical instantiation which alternates between integration of recent experience into the task distribution and meta-learning of the updated tasks. Repeating this procedure leads to iterative reorganization such that the curriculum adapts as the meta-learner's data distribution shifts. In particular, we show how discriminative clustering for visual representation can support trajectory-level task acquisition and exploration in domains with pixel observations, avoiding pitfalls of alternatives. In experiments on vision-based navigation and manipulation domains, we show that the algorithm allows for unsupervised meta-learning that transfers to downstream tasks specified by hand-crafted reward functions and serves as pre-training for more efficient supervised meta-learning of test task distributions.

Model-based reinforcement learning methods typically learn models for high-dimensional state spaces by aiming to reconstruct and predict the original observations. However, drawing inspiration from model-free reinforcement learning, we propose learning a latent dynamics model directly from rewards. In this work, we introduce a model-based planning framework which learns a latent reward prediction model and then plans in the latent state-space. The latent representation is learned exclusively from multi-step reward prediction which we show to be the only necessary information for successful planning. With this framework, we are able to benefit from the concise model-free representation, while still enjoying the data-efficiency of model-based algorithms. We demonstrate our framework in multi-pendulum and multi-cheetah environments where several pendulums or cheetahs are shown to the agent but only one of which produces rewards. In these environments, it is important for the agent to construct a concise latent representation to filter out irrelevant observations. We find that our method can successfully learn an accurate latent reward prediction model in the presence of the irrelevant information while existing model-based methods fail. Planning in the learned latent state-space shows strong performance and high sample efficiency over model-free and model-based baselines.

In this paper, we introduce ChainerRL, an open-source Deep Reinforcement Learning (DRL) library built using Python and the Chainer deep learning framework. ChainerRL implements a comprehensive set of DRL algorithms and techniques drawn from the state-of-the-art research in the field. To foster reproducible research, and for instructional purposes, ChainerRL provides scripts that closely replicate the original papers' experimental settings and reproduce published benchmark results for several algorithms. Lastly, ChainerRL offers a visualization tool that enables the qualitative inspection of trained agents. The ChainerRL source code can be found on GitHub: https://github.com/chainer/chainerrl .

There have been numerous attempts in explaining the general learning behaviours using model-based and model-free methods. While the model-based control is flexible yet computationally expensive in planning, the model-free control is quick but inflexible. The model-based control is therefore immune from reward devaluation and contingency degradation. Multiple arbitration schemes have been suggested to achieve the data efficiency and computational efficiency of model-based and model-free control respectively. In this context, we propose a quantitative 'value of information' based arbitration between both the controllers in order to establish a general computational framework for skill learning. The interacting model-based and model-free reinforcement learning processes are arbitrated using an uncertainty-based value of information. We further show that our algorithm performs better than Q-learning as well as Q-learning with experience replay.

We use Asynchronous Advantage Actor Critic (A3C) for implementing an AI agent in the controllers that optimize flow of traffic across a single intersection and then extend it to multiple intersections by considering a multi-agent setting. We explore three different methodologies to address the multi-agent problem - (1) use of asynchronous property of A3C to control multiple intersections using a single agent (2) utilise self/competitive play among independent agents across multiple intersections and (3) ingest a global reward function among agents to introduce cooperative behavior between intersections. We observe that (1) & (2) leads to a reduction in traffic congestion. Additionally the use of (3) with (1) & (2) led to a further reduction in congestion.

Multi-agent reinforcement learning (MARL) has long been a significant and everlasting research topic in both machine learning and control. With the recent development of (single-agent) deep RL, there is a resurgence of interests in developing new MARL algorithms, especially those that are backed by theoretical analysis. In this paper, we review some recent advances a sub-area of this topic: decentralized MARL with networked agents. Specifically, multiple agents perform sequential decision-making in a common environment, without the coordination of any central controller. Instead, the agents are allowed to exchange information with their neighbors over a communication network. Such a setting finds broad applications in the control and operation of robots, unmanned vehicles, mobile sensor networks, and smart grid. This review is built upon several our research endeavors in this direction, together with some progresses made by other researchers along the line. We hope this review to inspire the devotion of more research efforts to this exciting yet challenging area.

One of the primary factors behind the success of machine learning approaches in open world settings, such as image recognition and natural language processing, has been the ability of high-capacity deep neural network function approximators to learn generalizable models from large amounts of data. Deep reinforcement learning methods, however, require active online data collection, where the model actively interacts with its environment. This makes such methods hard to scale to complex real-world problems, where active data collection means that large datasets of experience must be collected for every experiment – this can be expensive and, for systems such as autonomous vehicles or robots, potentially unsafe. In a number of domains of practical interest, such as autonomous driving, robotics, and games, there exist plentiful amounts of previously collected interaction data which, consists of informative behaviours that are a rich source of prior information. Deep RL algorithms that can utilize such prior datasets will not only scale to real-world problems, but will also lead to solutions that generalize substantially better.

Note that any GIF compression artifacts in this animation are not present in the dataset itself. After collecting a diverse dataset, we experimentally investigate how it can be used to enable general skill learning that transfers to new environments. First, we pre-train visual dynamics models on a subset of data from RoboNet, and then fine-tune them to work in an unseen test environment using a small amount of new data. The constructed test environments (one of which is visualized below) all include different lab settings, new cameras and viewpoints, held-out robots, and novel objects purchased after data collection concluded. Example test environment constructed in a new lab, with a temporary uncalibrated camera, and a new Baxter robot.

When it comes to competitive games, AI systems have already shown they can easily mop the floor with the best humanity has to offer. But life in the real world isn't a zero sum game like poker or Starcraft and we need AI to work with us, not against us. That's why a research team from Facebook taught an AI how to play the cooperative card game Hanabi (the Japanese word for fireworks), to gain a better understanding of how humans think. Specifically, the Facebook team set out to instill upon its AI system the theory of mind. "Theory of mind is this idea of understanding the beliefs and intentions of other agents or other players or humans," Noam Brown, a researcher at Facebook AI, told Engadget.