We investigate the visual cross-embodiment imitation setting, in which agents learn policies from videos of other agents (such as humans) demonstrating the same task, but with stark differences in their embodiments -- shape, actions, end-effector dynamics, etc. In this work, we demonstrate that it is possible to automatically discover and learn vision-based reward functions from cross-embodiment demonstration videos that are robust to these differences. Specifically, we present a self-supervised method for Cross-embodiment Inverse Reinforcement Learning (XIRL) that leverages temporal cycle-consistency constraints to learn deep visual embeddings that capture task progression from offline videos of demonstrations across multiple expert agents, each performing the same task differently due to embodiment differences. Prior to our work, producing rewards from self-supervised embeddings has typically required alignment with a reference trajectory, which may be difficult to acquire. We show empirically that if the embeddings are aware of task-progress, simply taking the negative distance between the current state and goal state in the learned embedding space is useful as a reward for training policies with reinforcement learning. We find our learned reward function not only works for embodiments seen during training, but also generalizes to entirely new embodiments. We also find that XIRL policies are more sample efficient than baselines, and in some cases exceed the sample efficiency of the same agent trained with ground truth sparse rewards.

In this paper, we consider a transfer Reinforcement Learning (RL) problem in continuous state and action spaces, under unobserved contextual information. For example, the context can represent the mental view of the world that an expert agent has formed through past interactions with this world. We assume that this context is not accessible to a learner agent who can only observe the expert data. Then, our goal is to use the context-aware expert data to learn an optimal context-unaware policy for the learner using only a few new data samples. Such problems are typically solved using imitation learning that assumes that both the expert and learner agents have access to the same information. However, if the learner does not know the expert context, using the expert data alone will result in a biased learner policy and will require many new data samples to improve. To address this challenge, in this paper, we formulate the learning problem as a causal bound-constrained Multi-Armed-Bandit (MAB) problem. The arms of this MAB correspond to a set of basis policy functions that can be initialized in an unsupervised way using the expert data and represent the different expert behaviors affected by the unobserved context. On the other hand, the MAB constraints correspond to causal bounds on the accumulated rewards of these basis policy functions that we also compute from the expert data. The solution to this MAB allows the learner agent to select the best basis policy and improve it online. And the use of causal bounds reduces the exploration variance and, therefore, improves the learning rate. We provide numerical experiments on an autonomous driving example that show that our proposed transfer RL method improves the learner's policy faster compared to existing imitation learning methods and enjoys much lower variance during training.

We consider the problem of providing users of deep Reinforcement Learning (RL) based systems with a better understanding of when their output can be trusted. We offer an explainable artificial intelligence (XAI) framework that provides a three-fold explanation: a graphical depiction of the systems generalization and performance in the current game state, how well the agent would play in semantically similar environments, and a narrative explanation of what the graphical information implies. We created a user-interface for our XAI framework and evaluated its efficacy via a human-user experiment. The results demonstrate a statistically significant increase in user trust and acceptance of the AI system with explanation, versus the AI system without explanation.

Most of reinforcement learning algorithms optimize the discounted criterion which is beneficial to accelerate the convergence and reduce the variance of estimates. Although the discounted criterion is appropriate for certain tasks such as financial related problems, many engineering problems treat future rewards equally and prefer a long-run average criterion. In this paper, we study the reinforcement learning problem with the long-run average criterion. Firstly, we develop a unified trust region theory with discounted and average criteria. With the average criterion, a novel performance bound within the trust region is derived with the Perturbation Analysis (PA) theory. Secondly, we propose a practical algorithm named Average Policy Optimization (APO), which improves the value estimation with a novel technique named Average Value Constraint. To the best of our knowledge, our work is the first one to study the trust region approach with the average criterion and it complements the framework of reinforcement learning beyond the discounted criterion. Finally, experiments are conducted in the continuous control environment MuJoCo. In most tasks, APO performs better than the discounted PPO, which demonstrates the effectiveness of our approach.

Reinforcement learning competitions have formed the basis for standard research benchmarks, galvanized advances in the state-of-the-art, and shaped the direction of the field. Despite this, a majority of challenges suffer from the same fundamental problems: participant solutions to the posed challenge are usually domain-specific, biased to maximally exploit compute resources, and not guaranteed to be reproducible. In this paper, we present a new framework of competition design that promotes the development of algorithms that overcome these barriers. We propose four central mechanisms for achieving this end: submission retraining, domain randomization, desemantization through domain obfuscation, and the limitation of competition compute and environment-sample budget. To demonstrate the efficacy of this design, we proposed, organized, and ran the MineRL 2020 Competition on Sample-Efficient Reinforcement Learning. In this work, we describe the organizational outcomes of the competition and show that the resulting participant submissions are reproducible, non-specific to the competition environment, and sample/resource efficient, despite the difficult competition task.

A core issue with learning to optimize neural networks has been the lack of generalization to real world problems. To address this, we describe a system designed from a generalization-first perspective, learning to update optimizer hyperparameters instead of model parameters directly using novel features, actions, and a reward function. This system outperforms Adam at all neural network tasks including on modalities not seen during training. We achieve 2x speedups on ImageNet, and a 2.5x speedup on a language modeling task using over 5 orders of magnitude more compute than the training tasks.

The Reinforcement Learning (RL) paradigm involves a learning agent interacting with an unknown dynamical environment over multiple time steps. The learner receives a reward signal after each step which it uses to improve its performance over time. This formulation of RL has had significant empirical success in the recent past [Mni 15; Lev 16; Sil 17; Sen 20]. While this empirical success is encouraging, as RL starts to tackle a more wide-ranging class of consequential real-world problems, such as self-driving cars, supply chains, and medical care, a new set of challenges arise. Foremost among them is the lack of a well-specified reward signal associated with every state-action pair in many real-world settings. For example, consider a robot manipulation task where the robot must fold a pile of clothes. It is not clear how to design a useful reward signal that aids the robot to learn to complete this task. However, it is fairly easy to check whether the task was successfully completed (that is, whether the clothes were properly folded) and provide feedback at the end of the episode. This is a classical challenge but it is one that is often neglected in theoretical treatments of RL.

Learning from datasets without interaction with environments (Offline Learning) is an essential step to apply Reinforcement Learning (RL) algorithms in real-world scenarios. However, compared with the single-agent counterpart, offline multi-agent RL introduces more agents with the larger state and action space, which is more challenging but attracts little attention. We demonstrate current offline RL algorithms are ineffective in multi-agent systems due to the accumulated extrapolation error. In this paper, we propose a novel offline RL algorithm, named Implicit Constraint Q-learning (ICQ), which effectively alleviates the extrapolation error by only trusting the state-action pairs given in the dataset for value estimation. Moreover, we extend ICQ to multi-agent tasks by decomposing the joint-policy under the implicit constraint. Experimental results demonstrate that the extrapolation error is reduced to almost zero and insensitive to the number of agents. We further show that ICQ achieves the state-of-the-art performance in the challenging multi-agent offline tasks (StarCraft II).

We propose a novel framework for verifiable and compositional reinforcement learning (RL) in which a collection of RL sub-systems, each of which learns to accomplish a separate sub-task, are composed to achieve an overall task. The framework consists of a high-level model, represented as a parametric Markov decision process (pMDP) which is used to plan and to analyze compositions of sub-systems, and of the collection of low-level sub-systems themselves. By defining interfaces between the sub-systems, the framework enables automatic decompositons of task specifications, e.g., reach a target set of states with a probability of at least 0.95, into individual sub-task specifications, i.e. achieve the sub-system's exit conditions with at least some minimum probability, given that its entry conditions are met. This in turn allows for the independent training and testing of the sub-systems; if they each learn a policy satisfying the appropriate sub-task specification, then their composition is guaranteed to satisfy the overall task specification. Conversely, if the sub-task specifications cannot all be satisfied by the learned policies, we present a method, formulated as the problem of finding an optimal set of parameters in the pMDP, to automatically update the sub-task specifications to account for the observed shortcomings. The result is an iterative procedure for defining sub-task specifications, and for training the sub-systems to meet them. As an additional benefit, this procedure allows for particularly challenging or important components of an overall task to be determined automatically, and focused on, during training. Experimental results demonstrate the presented framework's novel capabilities.

We address the challenge of policy evaluation in real-world applications of reinforcement learning systems where the available historical data is limited due to ethical, practical, or security considerations. This constrained distribution of data samples often leads to biased policy evaluation estimates. To remedy this, we propose that instead of policy evaluation, one should perform policy comparison, i.e. to rank the policies of interest in terms of their value based on available historical data. In addition we present the Limited Data Estimator (LDE) as a simple method for evaluating and comparing policies from a small number of interactions with the environment. According to our theoretical analysis, the LDE is shown to be statistically reliable on policy comparison tasks under mild assumptions on the distribution of the historical data. Additionally, our numerical experiments compare the LDE to other policy evaluation methods on the task of policy ranking and demonstrate its advantage in various settings.