Reinforcement Learning (RL; Sutton and Barto, 2018; Mannor et al., 2022) studies online decision making problems in which an agent learns through experience within a dynamic environment, with the goal to minimize a loss function associated with the agent-environment interaction. Modern applications of RL such as robotics(Schulman et al., 2015; Lillicrap et al., 2015; Akkaya et al., 2019), game playing (Mnih et al., 2013; Silver et al., 2018) and autonomous driving (Kiran et al., 2021), almost invariably consist of large scale environments where function approximation techniques are necessary to allow the agent to generalize across different states. Furthermore, some form of agent robustness is usually required to cope with environment irregularities that cannot be faithfully represented by stochasticity assumptions (see e.g., Dulac-Arnold et al., 2021). Theoretical foundations for RL with function approximation (e.g., Jiang et al., 2017; Yang and Wang, 2019; Jin et al., 2020b; Agarwal et al., 2020) have been steadily coming into fruition.

Many instances of similar or almost-identical industrial machines or tools are often deployed at once, or in quick succession. For instance, a particular model of air compressor may be installed at hundreds of customers. Because these tools perform distinct but highly similar tasks, it is interesting to be able to quickly produce a high-quality controller for machine $N+1$ given the controllers already produced for machines $1..N$. This is even more important when the controllers are learned through Reinforcement Learning, as training takes time, energy and other resources. In this paper, we apply Policy Intersection, a Policy Shaping method, to help a Reinforcement Learning agent learn to solve a new variant of a compressors control problem faster, by transferring knowledge from several previously learned controllers. We show that our approach outperforms loading an old controller, and significantly improves performance in the long run.

Deep reinforcement learning has recently been applied to a variety of robotics applications, but learning locomotion for robots with unconventional configurations is still limited. Prior work has shown that, despite the simple modeling of articulated swimmer robots, such systems struggle to find effective gaits using reinforcement learning due to the heterogeneity of the search space. In this work, we leverage insight from geometric models of these robots in order to focus on promising regions of the space and guide the learning process. We demonstrate that our augmented learning technique is able to produce gaits for different learning goals for swimmer robots in both low and high Reynolds number fluids.

Hindsight goal relabeling has become a foundational technique in multi-goal reinforcement learning (RL). The essential idea is that any trajectory can be seen as a sub-optimal demonstration for reaching its final state. Intuitively, learning from those arbitrary demonstrations can be seen as a form of imitation learning (IL). However, the connection between hindsight goal relabeling and imitation learning is not well understood. In this paper, we propose a novel framework to understand hindsight goal relabeling from a divergence minimization perspective. Recasting the goal reaching problem in the IL framework not only allows us to derive several existing methods from first principles, but also provides us with the tools from IL to improve goal reaching algorithms. Experimentally, we find that under hindsight relabeling, Q-learning outperforms behavioral cloning (BC). Yet, a vanilla combination of both hurts performance. Concretely, we see that the BC loss only helps when selectively applied to actions that get the agent closer to the goal according to the Q-function. Our framework also explains the puzzling phenomenon wherein a reward of (-1, 0) results in significantly better performance than a (0, 1) reward for goal reaching.

Training large neural networks is known to be time-consuming, with the learning duration taking days or even weeks. To address this problem, large-batch optimization was introduced. This approach demonstrated that scaling mini-batch sizes with appropriate learning rate adjustments can speed up the training process by orders of magnitude. While long training time was not typically a major issue for model-free deep offline RL algorithms, recently introduced Q-ensemble methods achieving state-of-the-art performance made this issue more relevant, notably extending the training duration. In this work, we demonstrate how this class of methods can benefit from large-batch optimization, which is commonly overlooked by the deep offline RL community. We show that scaling the mini-batch size and naively adjusting the learning rate allows for (1) a reduced size of the Q-ensemble, (2) stronger penalization of out-of-distribution actions, and (3) improved convergence time, effectively shortening training duration by 3-4x times on average.

In safety-critical decision-making scenarios being able to identify worst-case outcomes, or dead-ends is crucial in order to develop safe and reliable policies in practice. These situations are typically rife with uncertainty due to unknown or stochastic characteristics of the environment as well as limited offline training data. As a result, the value of a decision at any time point should be based on the distribution of its anticipated effects. We propose a framework to identify worst-case decision points, by explicitly estimating distributions of the expected return of a decision. These estimates enable earlier indication of dead-ends in a manner that is tunable based on the risk tolerance of the designed task. We demonstrate the utility of Distributional Dead-end Discovery (DistDeD) in a toy domain as well as when assessing the risk of severely ill patients in the intensive care unit reaching a point where death is unavoidable. We find that DistDeD significantly improves over prior discovery approaches, providing indications of the risk 10 hours earlier on average as well as increasing detection by 20%.

Therefore, we require methods that can guarantee robust and safe performance under general forms of environment uncertainty. Robustness and safety are critical for the trustworthy Unfortunately, popular approaches to robustness deployment of deep reinforcement learning in deep RL consider very structured forms of uncertainty in real-world decision making applications. In in order to facilitate efficient implementations. Adversarial particular, we require algorithms that can guarantee methods implement a specific type of perturbation, such robust, safe performance in the presence of as the application of a physical force (Pinto et al., 2017) general environment disturbances, while making or a change in the action that is deployed (Tessler et al., limited assumptions on the data collection process 2019a). Parametric approaches, on the other hand, consider during training. In this work, we propose a safe reinforcement robustness with respect to environment characteristics that learning framework with robustness can be altered in a simulator (Rajeswaran et al., 2017; Peng guarantees through the use of an optimal transport et al., 2018; Mankowitz et al., 2020). When we lack domain cost uncertainty set. We provide an efficient, knowledge on the structure of potential disturbances, these theoretically supported implementation based on techniques may not guarantee robustness and safety. Optimal Transport Perturbations, which can be applied in a completely offline fashion using only Another drawback of existing approaches is their need to data collected in a nominal training environment.

Deep Q-learning based algorithms have been applied successfully in many decision making problems, while their theoretical foundations are not as well understood. In this paper, we study a Fitted Q-Iteration with two-layer ReLU neural network parameterization, and find the sample complexity guarantees for the algorithm. Our approach estimates the Q-function in each iteration using a convex optimization problem. We show that this approach achieves a sample complexity of $\tilde{\mathcal{O}}(1/\epsilon^{2})$, which is order-optimal. This result holds for a countable state-spaces and does not require any assumptions such as a linear or low rank structure on the MDP.

There are many frameworks for deep generative modeling, each often presented with their own specific training algorithms and inference methods. Here, we demonstrate the connections between existing deep generative models and the recently introduced GFlowNet framework, a probabilistic inference machine which treats sampling as a decision-making process. This analysis sheds light on their overlapping traits and provides a unifying viewpoint through the lens of learning with Markovian trajectories. Our framework provides a means for unifying training and inference algorithms, and provides a route to shine a unifying light over many generative models. Beyond this, we provide a practical and experimentally verified recipe for improving generative modeling with insights from the GFlowNet perspective.

Decentralized Actor-Critic (AC) algorithms have been widely utilized for multi-agent reinforcement learning (MARL) and have achieved remarkable success. Apart from its empirical success, the theoretical convergence property of decentralized AC algorithms is largely unexplored. Most of the existing finite-time convergence results are derived based on either double-loop update or two-timescale step sizes rule, and this is the case even for centralized AC algorithm under a single-agent setting. In practice, the \emph{single-timescale} update is widely utilized, where actor and critic are updated in an alternating manner with step sizes being of the same order. In this work, we study a decentralized \emph{single-timescale} AC algorithm.Theoretically, using linear approximation for value and reward estimation, we show that the algorithm has sample complexity of $\tilde{\mathcal{O}}(\varepsilon^{-2})$ under Markovian sampling, which matches the optimal complexity with a double-loop implementation (here, $\tilde{\mathcal{O}}$ hides a logarithmic term). When we reduce to the single-agent setting, our result yields new sample complexity for centralized AC using a single-timescale update scheme. The central to establishing our complexity results is \emph{the hidden smoothness of the optimal critic variable} we revealed. We also provide a local action privacy-preserving version of our algorithm and its analysis. Finally, we conduct experiments to show the superiority of our algorithm over the existing decentralized AC algorithms.