Offline safe reinforcement learning (RL) has emerged as a promising approach for learning safe behaviors without engaging in risky online interactions with the environment. Most existing methods in offline safe RL rely on cost constraints at each time step (derived from global cost constraints) and this can result in either overly conservative policies or violation of safety constraints. In this paper, we propose to learn a policy that generates desirable trajectories and avoids undesirable trajectories. To be specific, we first partition the pre-collected dataset of state-action trajectories into desirable and undesirable subsets. Intuitively, the desirable set contains high reward and safe trajectories, and undesirable set contains unsafe trajectories and low-reward safe trajectories. Second, we learn a policy that generates desirable trajectories and avoids undesirable trajectories, where (un)desirability scores are provided by a classifier learnt from the dataset of desirable and undesirable trajectories. This approach bypasses the computational complexity and stability issues of a min-max objective that is employed in existing methods. Theoretically, we also show our approach's strong connections to existing learning paradigms involving human feedback. Finally, we extensively evaluate our method using the DSRL benchmark for offline safe RL. Empirically, our method outperforms competitive baselines, achieving higher rewards and better constraint satisfaction across a wide variety of benchmark tasks.

Reinforcement learning (RL) can be formulated as a sequence modeling problem, where models predict future actions based on historical state-action-reward sequences. Current approaches typically require long trajectory sequences to model the environment in offline RL settings. However, these models tend to over-rely on memorizing long-term representations, which impairs their ability to effectively attribute importance to trajectories and learned representations based on task-specific relevance. In this work, we introduce AdaCred, a novel approach that represents trajectories as causal graphs built from short-term action-reward-state sequences. Our model adaptively learns control policy by crediting and pruning low-importance representations, retaining only those most relevant for the downstream task. Our experiments demonstrate that AdaCred-based policies require shorter trajectory sequences and consistently outperform conventional methods in both offline reinforcement learning and imitation learning environments.

The ability to estimate temporal relationships is critical for both animals and artificial agents. Cognitive science and neuroscience provide remarkable insights into behavioral and neural aspects of temporal credit assignment. In particular, scale invariance of learning dynamics, observed in behavior and supported by neural data, is one of the key principles that governs animal perception: proportional rescaling of temporal relationships does not alter the overall learning efficiency. Here we integrate a computational neuroscience model of scale invariant memory into deep reinforcement learning (RL) agents. We first provide a theoretical analysis and then demonstrate through experiments that such agents can learn robustly across a wide range of temporal scales, unlike agents built with commonly used recurrent memory architectures such as LSTM. This result illustrates that incorporating computational principles from neuroscience and cognitive science into deep neural networks can enhance adaptability to complex temporal dynamics, mirroring some of the core properties of human learning.

Recent advancements in robot learning have used imitation learning with large models and extensive demonstrations to develop effective policies. However, these models are often limited by the quantity, quality, and diversity of demonstrations. This paper explores improving offline-trained imitation learning models through online interactions with the environment. We introduce Policy Decorator, which uses a model-agnostic residual policy to refine large imitation learning models during online interactions. By implementing controlled exploration strategies, Policy Decorator enables stable, sample-efficient online learning. Our evaluation spans eight tasks across two benchmarks-ManiSkill and Adroit-and involves two state-of-the-art imitation learning models (Behavior Transformer and Diffusion Policy). The results show Policy Decorator effectively improves the offline-trained policies and preserves the smooth motion of imitation learning models, avoiding the erratic behaviors of pure RL policies. See our project page (https://policydecorator.github.io) for videos.

Airborne Wind Energy (AWE) is an emerging technology designed to harness the power of high-altitude winds, offering a solution to several limitations of conventional wind turbines. AWE is based on flying devices (usually gliders or kites) that, tethered to a ground station and driven by the wind, convert its mechanical energy into electrical energy by means of a generator. Such systems are usually controlled by manoeuvering the kite so as to follow a predefined path prescribed by optimal control techniques, such as model-predictive control. These methods are strongly dependent on the specific model at use and difficult to generalize, especially in unpredictable conditions such as the turbulent atmospheric boundary layer. Our aim is to explore the possibility of replacing these techniques with an approach based on Reinforcement Learning (RL). Unlike traditional methods, RL does not require a predefined model, making it robust to variability and uncertainty. Our experimental results in complex simulated environments demonstrate that AWE agents trained with RL can effectively extract energy from turbulent flows, relying on minimal local information about the kite orientation and speed relative to the wind.

Learning policies from high-dimensional visual inputs, such as pixels and point clouds, is crucial in various applications. Visual reinforcement learning is a promising approach that directly trains policies from visual observations, although it faces challenges in sample efficiency and computational costs. This study conducts an empirical comparison of State-to-Visual DAgger, a two-stage framework that initially trains a state policy before adopting online imitation to learn a visual policy, and Visual RL across a diverse set of tasks. We evaluate both methods across 16 tasks from three benchmarks, focusing on their asymptotic performance, sample efficiency, and computational costs. Surprisingly, our findings reveal that State-to-Visual DAgger does not universally outperform Visual RL but shows significant advantages in challenging tasks, offering more consistent performance. In contrast, its benefits in sample efficiency are less pronounced, although it often reduces the overall wall-clock time required for training. Based on our findings, we provide recommendations for practitioners and hope that our results contribute valuable perspectives for future research in visual policy learning.

This paper investigates the use of multi-agent reinforcement learning (MARL) to address distributed channel access in wireless local area networks. In particular, we consider the challenging yet more practical case where the agents heterogeneously adopt value-based or policy-based reinforcement learning algorithms to train the model. We propose a heterogeneous MARL training framework, named QPMIX, which adopts a centralized training with distributed execution paradigm to enable heterogeneous agents to collaborate. Moreover, we theoretically prove the convergence of the proposed heterogeneous MARL method when using the linear value function approximation. Our method maximizes the network throughput and ensures fairness among stations, therefore, enhancing the overall network performance. Simulation results demonstrate that the proposed QPMIX algorithm improves throughput, mean delay, delay jitter, and collision rates compared with conventional carrier-sense multiple access with collision avoidance in the saturated traffic scenario. Furthermore, the QPMIX is shown to be robust in unsaturated and delay-sensitive traffic scenarios, and promotes cooperation among heterogeneous agents.

Large-scale crises, including wars and pandemics, have repeatedly shaped human history, and their simultaneous occurrence presents profound challenges to societies. Understanding the dynamics of epidemic spread during warfare is essential for developing effective containment strategies in complex conflict zones. While research has explored epidemic models in various settings, the impact of warfare on epidemic dynamics remains underexplored. In this study, we proposed a novel mathematical model that integrates the epidemiological SIR (susceptible-infected-recovered) model with the war dynamics Lanchester model to explore the dual influence of war and pandemic on a population's mortality. Moreover, we consider a dual-use military and civil healthcare system that aims to reduce the overall mortality rate which can use different administration policies. Using an agent-based simulation to generate in silico data, we trained a deep reinforcement learning model for healthcare administration policy and conducted an intensive investigation on its performance. Our results show that a pandemic during war conduces chaotic dynamics where the healthcare system should either prioritize war-injured soldiers or pandemic-infected civilians based on the immediate amount of mortality from each option, ignoring long-term objectives. Our findings highlight the importance of integrating conflict-related factors into epidemic modeling to enhance preparedness and response strategies in conflict-affected areas.

Realtime environments change even as agents perform action inference and learning, thus requiring high interaction frequencies to effectively minimize regret. However, recent advances in machine learning involve larger neural networks with longer inference times, raising questions about their applicability in realtime systems where reaction time is crucial. We present an analysis of lower bounds on regret in realtime reinforcement learning (RL) environments to show that minimizing long-term regret is generally impossible within the typical sequential interaction and learning paradigm, but often becomes possible when sufficient asynchronous compute is available. We propose novel algorithms for staggering asynchronous inference processes to ensure that actions are taken at consistent time intervals, and demonstrate that use of models with high action inference times is only constrained by the environment's effective stochasticity over the inference horizon, and not by action frequency. Our analysis shows that the number of inference processes needed scales linearly with increasing inference times while enabling use of models that are multiple orders of magnitude larger than existing approaches when learning from a realtime simulation of Game Boy games such as Pok\'emon and Tetris.

Over the past few years, Reinforcement Learning has shown to have the capacity to train Deep Neural Networks to perform complex tasks. This paper investigates the use of a Deep Reinforcement Learning algorithm, Twin Delayed Deep Deterministic Policy Gradient, to learn a policy to fly a quadcopter through a First Person View(FPV) drone racing gate. BattleDrones is an autonomous drone racing competition held by Virginia Tech. Teams must design a controller to navigate a quadcopter through a course consisting of multiple gates as part of the competition. The quadcopter is outfitted with a camera that is used to identify an AprilTag [1], a fiducial marker, on the gates.