Transformers are neural network models that utilize multiple layers of self-attention heads and have exhibited enormous potential in natural language processing tasks. Meanwhile, there have been efforts to adapt transformers to visual tasks of machine learning, including Vision Transformers and Swin Transformers. Although some researchers use Vision Transformers for reinforcement learning tasks, their experiments remain at a small scale due to the high computational cost. Experiments conducted at a large scale, on the other hand, have to rely on techniques to cut the costs of Vision Transformers, which also yield inferior results. To address this challenge, this article presents the first online reinforcement learning scheme that is based on Swin Transformers: Swin DQN. Swin Transformers are promising as a backbone in neural networks by splitting groups of image pixels into small patches and applying local self-attention operations inside the (shifted) windows of fixed sizes. They have demonstrated state-of-the-art performances in benchmarks. In contrast to existing research, our novel approach is reducing the computational costs, as well as significantly improving the performance. We demonstrate the superior performance with experiments on 49 games in the Arcade Learning Environment. The results show that our approach, using Swin Transformers with Double DQN, achieves significantly higher maximal evaluation scores than the baseline method in 45 of all the 49 games ~92%, and higher mean evaluation scores than the baseline method in 40 of all the 49 games ~82%.

The field of Continual Learning (CL) seeks to develop algorithms that accumulate knowledge and skills over time through interaction with non-stationary environments. In practice, a plethora of evaluation procedures (settings) and algorithmic solutions (methods) exist, each with their own potentially disjoint set of assumptions. This variety makes measuring progress in CL difficult. We propose a taxonomy of settings, where each setting is described as a set of assumptions. A tree-shaped hierarchy emerges from this view, where more general settings become the parents of those with more restrictive assumptions. This makes it possible to use inheritance to share and reuse research, as developing a method for a given setting also makes it directly applicable onto any of its children. We instantiate this idea as a publicly available software framework called Sequoia, which features a wide variety of settings from both the Continual Supervised Learning (CSL) and Continual Reinforcement Learning (CRL) domains. Sequoia also includes a growing suite of methods which are easy to extend and customize, in addition to more specialized methods from external libraries. We hope that this new paradigm and its first implementation can help unify and accelerate research in CL. You can help us grow the tree by visiting www.github.com/lebrice/Sequoia.

Many real-world settings involve costs for performing actions; transaction costs in financial systems and fuel costs being common examples. In these settings, performing actions at each time step quickly accumulates costs leading to vastly suboptimal outcomes. Additionally, repeatedly acting produces wear and tear and ultimately, damage. Determining \textit{when to act} is crucial for achieving successful outcomes and yet, the challenge of efficiently \textit{learning} to behave optimally when actions incur minimally bounded costs remains unresolved. In this paper, we introduce a reinforcement learning (RL) framework named \textbf{L}earnable \textbf{I}mpulse \textbf{C}ontrol \textbf{R}einforcement \textbf{A}lgorithm (LICRA), for learning to optimally select both when to act and which actions to take when actions incur costs. At the core of LICRA is a nested structure that combines RL and a form of policy known as \textit{impulse control} which learns to maximise objectives when actions incur costs. We prove that LICRA, which seamlessly adopts any RL method, converges to policies that optimally select when to perform actions and their optimal magnitudes. We then augment LICRA to handle problems in which the agent can perform at most $k<\infty$ actions and more generally, faces a budget constraint. We show LICRA learns the optimal value function and ensures budget constraints are satisfied almost surely. We demonstrate empirically LICRA's superior performance against benchmark RL methods in OpenAI gym's \textit{Lunar Lander} and in \textit{Highway} environments and a variant of the Merton portfolio problem within finance.

Hierarchical reinforcement learning (HRL; Barto & Mahadevan, 2003) frameworks, such as options (Sutton et al., Reward machines (RMs) are a recent formalism 1999), have been used to exploit RMs by learning policies for representing the reward function of a reinforcement at two levels of abstraction: (i) select a formula (i.e., subgoal) learning task through a finite-state machine from a given RM state, and (ii) select an action to whose edges encode subgoals of the task using (eventually) satisfy the chosen formula (Toro Icarte et al., high-level events. The structure of RMs enables 2018; Furelos-Blanco et al., 2021). The subtask decomposition the decomposition of a task into simpler and independently powered by HRL enables learning at multiple scales solvable subtasks that help tackle longhorizon simultaneously, and eases the handling of long-horizon and and/or sparse reward tasks. We propose sparse reward tasks. In addition, several works have considered a formalism for further abstracting the subtask the problem of learning the RMs themselves from structure by endowing an RM with the ability to interaction (e.g., Toro Icarte et al., 2019; Xu et al., 2020; call other RMs, thus composing a hierarchy of Furelos-Blanco et al., 2021; Hasanbeig et al., 2021).

In fully cooperative multi-agent reinforcement learning (MARL) settings, environments are highly stochastic due to the partial observability of each agent and the continuously changing policies of other agents. To address the above issues, we proposed a unified framework, called DFAC, for integrating distributional RL with value function factorization methods. This framework generalizes expected value function factorization methods to enable the factorization of return distributions. To validate DFAC, we first demonstrate its ability to factorize the value functions of a simple matrix game with stochastic rewards. Then, we perform experiments on all Super Hard maps of the StarCraft Multi-Agent Challenge and six self-designed Ultra Hard maps, showing that DFAC is able to outperform a number of baselines.

Reinforcement learning agents may sometimes develop habits that are effective only when specific policies are followed. After an initial exploration phase in which agents try out different actions, they eventually converge toward a particular policy. When this occurs, the distribution of state-action trajectories becomes narrower, and agents start experiencing the same transitions again and again. At this point, spurious correlations may arise. Agents may then pick up on these correlations and learn state representations that do not generalize beyond the agent's trajectory distribution. In this paper, we provide a mathematical characterization of this phenomenon, which we refer to as policy confounding, and show, through a series of examples, when and how it occurs in practice.

In meta reinforcement learning (meta RL), an agent seeks a Bayes-optimal policy -- the optimal policy when facing an unknown task that is sampled from some known task distribution. Previous approaches tackled this problem by inferring a belief over task parameters, using variational inference methods. Motivated by recent successes of contrastive learning approaches in RL, such as contrastive predictive coding (CPC), we investigate whether contrastive methods can be used for learning Bayes-optimal behavior. We begin by proving that representations learned by CPC are indeed sufficient for Bayes optimality. Based on this observation, we propose a simple meta RL algorithm that uses CPC in lieu of variational belief inference. Our method, ContraBAR, achieves comparable performance to state-of-the-art in domains with state-based observation and circumvents the computational toll of future observation reconstruction, enabling learning in domains with image-based observations. It can also be combined with image augmentations for domain randomization and used seamlessly in both online and offline meta RL settings.

We study finite episodic Markov decision processes incorporating dynamic risk measures to capture risk sensitivity. To this end, we present two model-based algorithms applied to \emph{Lipschitz} dynamic risk measures, a wide range of risk measures that subsumes spectral risk measure, optimized certainty equivalent, distortion risk measures among others. We establish both regret upper bounds and lower bounds. Notably, our upper bounds demonstrate optimal dependencies on the number of actions and episodes, while reflecting the inherent trade-off between risk sensitivity and sample complexity. Additionally, we substantiate our theoretical results through numerical experiments.

Exploration remains a key challenge in deep reinforcement learning (RL). Optimism in the face of uncertainty is a well-known heuristic with theoretical guarantees in the tabular setting, but how best to translate the principle to deep reinforcement learning, which involves online stochastic gradients and deep network function approximators, is not fully understood. In this paper we propose a new, differentiable optimistic objective that when optimized yields a policy that provably explores efficiently, with guarantees even under function approximation. Our new objective is a zero-sum two-player game derived from endowing the agent with an epistemic-risk-seeking utility function, which converts uncertainty into value and encourages the agent to explore uncertain states. We show that the solution to this game minimizes an upper bound on the regret, with the 'players' each attempting to minimize one component of a particular regret decomposition. We derive a new model-free algorithm which we call 'epistemic-risk-seeking actor-critic' (ERSAC), which is simply an application of simultaneous stochastic gradient ascent-descent to the game. Finally, we discuss a recipe for incorporating off-policy data and show that combining the risk-seeking objective with replay data yields a double benefit in terms of statistical efficiency. We conclude with some results showing good performance of a deep RL agent using the technique on the challenging 'DeepSea' environment, showing significant performance improvements even over other efficient exploration techniques, as well as improved performance on the Atari benchmark.

Traffic simulation provides interactive data for the optimization of traffic control policies. However, existing traffic simulators are limited by their lack of scalability and shortage in input data, which prevents them from generating interactive data from traffic simulation in the scenarios of real large-scale city road networks. In this paper, we present \textbf{C}ity \textbf{B}rain \textbf{Lab}, a toolkit for scalable traffic simulation. CBLab consists of three components: CBEngine, CBData, and CBScenario. CBEngine is a highly efficient simulator supporting large-scale traffic simulation. CBData includes a traffic dataset with road network data of 100 cities all around the world. We also develop a pipeline to conduct a one-click transformation from raw road networks to input data of our traffic simulation. Combining CBEngine and CBData allows researchers to run scalable traffic simulations in the road network of real large-scale cities. Based on that, CBScenario implements an interactive environment and a benchmark for two scenarios of traffic control policies respectively, with which traffic control policies adaptable for large-scale urban traffic can be trained and tuned. To the best of our knowledge, CBLab is the first infrastructure supporting traffic control policy optimization in large-scale urban scenarios. CBLab has supported the City Brain Challenge @ KDD CUP 2021. The project is available on GitHub:~\url{https://github.com/CityBrainLab/CityBrainLab.git}.