Partial observability of the underlying states generally presents significant challenges for reinforcement learning (RL). In practice, certain \emph{privileged information}, e.g., the access to states from simulators, has been exploited in training and has achieved prominent empirical successes. To better understand the benefits of privileged information, we revisit and examine several simple and practically used paradigms in this setting. Specifically, we first formalize the empirical paradigm of \emph{expert distillation} (also known as \emph{teacher-student} learning), demonstrating its pitfall in finding near-optimal policies. We then identify a condition of the partially observable environment, the \emph{deterministic filter condition}, under which expert distillation achieves sample and computational complexities that are \emph{both} polynomial. Furthermore, we investigate another useful empirical paradigm of \emph{asymmetric actor-critic}, and focus on the more challenging setting of observable partially observable Markov decision processes. We develop a belief-weighted asymmetric actor-critic algorithm with polynomial sample and quasi-polynomial computational complexities, in which one key component is a new provable oracle for learning belief states that preserve \emph{filter stability} under a misspecified model, which may be of independent interest. Finally, we also investigate the provable efficiency of partially observable multi-agent RL (MARL) with privileged information. We develop algorithms featuring \emph{centralized-training-with-decentralized-execution}, a popular framework in empirical MARL, with polynomial sample and (quasi-)polynomial computational complexities in both paradigms above. Compared with a few recent related theoretical studies, our focus is on understanding practically inspired algorithmic paradigms, without computationally intractable oracles.

This paper proposes an online environment poisoning algorithm tailored for reinforcement learning agents operating in a black-box setting, where an adversary deliberately manipulates training data to lead the agent toward a mischievous policy. In contrast to prior studies that primarily investigate white-box settings, we focus on a scenario characterized by \textit{unknown} environment dynamics to the attacker and a \textit{flexible} reinforcement learning algorithm employed by the targeted agent. We first propose an attack scheme that is capable of poisoning the reward functions and state transitions. The poisoning task is formalized as a constrained optimization problem, following the framework of \cite{ma2019policy}. Given the transition probabilities are unknown to the attacker in a black-box environment, we apply a stochastic gradient descent algorithm, where the exact gradients are approximated using sample-based estimates. A penalty-based method along with a bilevel reformulation is then employed to transform the problem into an unconstrained counterpart and to circumvent the double-sampling issue. The algorithm's effectiveness is validated through a maze environment.

Many continuous control problems can be formulated as sparse-reward reinforcement learning (RL) tasks. In principle, online RL methods can automatically explore the state space to solve each new task. However, discovering sequences of actions that lead to a non-zero reward becomes exponentially more difficult as the task horizon increases. Manually shaping rewards can accelerate learning for a fixed task, but it is an arduous process that must be repeated for each new environment. We introduce a systematic reward-shaping framework that distills the information contained in 1) a task-agnostic prior data set and 2) a small number of task-specific expert demonstrations, and then uses these priors to synthesize dense dynamics-aware rewards for the given task. This supervision substantially accelerates learning in our experiments, and we provide analysis demonstrating how the approach can effectively guide online learning agents to faraway goals. Keywords: Imitation Learning, Learning from Demonstrations, Reward Shaping.

Truck platooning technology enables a group of trucks to travel closely together, with which the platoon can save fuel, improve traffic flow efficiency, and improve safety. In this paper, we consider the platoon coordination problem in a large-scale transportation network, to promote cooperation among trucks and optimize the overall efficiency. Involving the regulation of both speed and departure times at hubs, we formulate the coordination problem as a complicated dynamic stochastic integer programming under network and information constraints. To get an autonomous, distributed, and robust platoon coordination policy, we formulate the problem into a model of the Decentralized-Partial Observable Markov Decision Process. Then, we propose a Multi-Agent Deep Reinforcement Learning framework named Trcuk Attention-QMIX (TA-QMIX) to train an efficient online decision policy. TA-QMIX utilizes the attention mechanism to enhance the representation of truck fuel gains and delay times, and provides explicit truck cooperation information during the training process, promoting trucks' willingness to cooperate. The training framework adopts centralized training and distributed execution, thus training a policy for trucks to make decisions online using only nearby information. Hence, the policy can be autonomously executed on a large-scale network. Finally, we perform comparison experiments and ablation experiments in the transportation network of the Yangtze River Delta region in China to verify the effectiveness of the proposed framework. In a repeated comparative experiment with 5,000 trucks, our method average saves 19.17\% of fuel with an average delay of only 9.57 minutes per truck and a decision time of 0.001 seconds.

In this paper, we consider a point-to-point integrated sensing and communication (ISAC) system, where a transmitter conveys a message to a receiver over a channel with memory and simultaneously estimates the state of the channel through the backscattered signals from the emitted waveform. Using Massey's concept of directed information for channels with memory, we formulate the capacity-distortion tradeoff for the ISAC problem when sensing is performed in an online fashion. Optimizing the transmit waveform for this system to simultaneously achieve good communication and sensing performance is a complicated task, and thus we propose a deep reinforcement learning (RL) approach to find a solution. The proposed approach enables the agent to optimize the ISAC performance by learning a reward that reflects the difference between the communication gain and the sensing loss. Since the state-space in our RL model is \`a priori unbounded, we employ deep deterministic policy gradient algorithm (DDPG). Our numerical results suggest a significant performance improvement when one considers unbounded state-space as opposed to a simpler RL problem with reduced state-space. In the extreme case of degenerate state-space only memoryless signaling strategies are possible. Our results thus emphasize the necessity of well exploiting the memory inherent in ISAC systems.

Drone Visual Active Tracking aims to autonomously follow a target object by controlling the motion system based on visual observations, providing a more practical solution for effective tracking in dynamic environments. However, accurate Drone Visual Active Tracking using reinforcement learning remains challenging due to the absence of a unified benchmark, the complexity of open-world environments with frequent interference, and the diverse motion behavior of dynamic targets. To address these issues, we propose a unified cross-scene cross-domain benchmark for open-world drone active tracking called DAT. The DAT benchmark provides 24 visually complex environments to assess the algorithms' cross-scene and cross-domain generalization abilities, and high-fidelity modeling of realistic robot dynamics. Additionally, we propose a reinforcement learning-based drone tracking method called R-VAT, which aims to improve the performance of drone tracking targets in complex scenarios. Specifically, inspired by curriculum learning, we introduce a Curriculum-Based Training strategy that progressively enhances the agent tracking performance in vast environments with complex interference. We design a goal-centered reward function to provide precise feedback to the drone agent, preventing targets farther from the center of view from receiving higher rewards than closer ones. This allows the drone to adapt to the diverse motion behavior of open-world targets. Experiments demonstrate that the R-VAT has about 400% improvement over the SOTA method in terms of the cumulative reward metric.

InvestESG is a novel multi-agent reinforcement learning (MARL) benchmark designed to study the impact of Environmental, Social, and Governance (ESG) disclosure mandates on corporate climate investments. Supported by both PyTorch and JAX implementation, the benchmark models an intertemporal social dilemma where companies balance short-term profit losses from climate mitigation efforts and long-term benefits from reducing climate risk, while ESG-conscious investors attempt to influence corporate behavior through their investment decisions, in a scalable and hardware-accelerated manner. Companies allocate capital across mitigation, greenwashing, and resilience, with varying strategies influencing climate outcomes and investor preferences. Our experiments show that without ESG-conscious investors with sufficient capital, corporate mitigation efforts remain limited under the disclosure mandate. However, when a critical mass of investors prioritizes ESG, corporate cooperation increases, which in turn reduces climate risks and enhances long-term financial stability. Additionally, providing more information about global climate risks encourages companies to invest more in mitigation, even without investor involvement. Our findings align with empirical research using real-world data, highlighting MARL's potential to inform policy by providing insights into large-scale socio-economic challenges through efficient testing of alternative policy and market designs.

While Deep Reinforcement Learning (DRL) has achieved remarkable success across various domains, it remains vulnerable to occasional catastrophic failures without additional safeguards. An effective solution to prevent these failures is to use a shield that validates and adjusts the agent's actions to ensure compliance with a provided set of safety specifications. For real-world robotic domains, it is essential to define safety specifications over continuous state and action spaces to accurately account for system dynamics and compute new actions that minimally deviate from the agent's original decision. In this paper, we present the first shielding approach specifically designed to ensure the satisfaction of safety requirements in continuous state and action spaces, making it suitable for practical robotic applications. Our method builds upon realizability, an essential property that confirms the shield will always be able to generate a safe action for any state in the environment. We formally prove that realizability can be verified for stateful shields, enabling the incorporation of non-Markovian safety requirements, such as loop avoidance. Finally, we demonstrate the effectiveness of our approach in ensuring safety without compromising the policy's success rate by applying it to a navigation problem and a multi-agent particle environment Keywords: Shielding, Reinforcement Learning, Safety, Robotics

Combinatorial online learning is a fundamental task to decide the optimal combination of base arms in sequential interactions with systems providing uncertain rewards, which is applicable to diverse domains such as robotics, social advertising, network routing and recommendation systems. In real-world scenarios, we often observe rising rewards, where the selection of a base arm not only provides an instantaneous reward but also contributes to the enhancement of future rewards, {\it e.g.}, robots enhancing proficiency through practice and social influence strengthening in the history of successful recommendations. To address this, we introduce the problem of combinatorial rising bandit to minimize policy regret and propose a provably efficient algorithm, called Combinatorial Rising Upper Confidence Bound (CRUCB), of which regret upper bound is close to a regret lower bound. To the best of our knowledge, previous studies do not provide a sub-linear regret lower bound, making it impossible to assess the efficiency of their algorithms. However, we provide the sub-linear regret lower bound for combinatorial rising bandit and show that CRUCB is provably efficient by showing that the regret upper bound is close to the regret lower bound. In addition, we empirically demonstrate the effectiveness and superiority of CRUCB not only in synthetic environments but also in realistic applications of deep reinforcement learning.

Reinforcement Learning (RL) has become a popular learning framework to solve sequential decision making problems in unknown environments, and has achieved tremendous success in a wide array of domains such as playing the game of Go (Silver et al., 2016), robotic control (Kober et al., 2013), and autonomous driving (Kiran et al., 2022; Lin et al., 2023). A critical feature of most real-world systems is their uncertain nature, and consequently RL has emerged as a powerful tool for learning optimal policies for multi-agent systems to operate in unknown environments (Kim & Giannakis, 2017; Zhang et al., 2021; Lin et al., 2024; Anand & Qu, 2024). While the early literature on RL predominantly focused on the single-agent setting, multi-agent reinforcement learning (MARL) has also recently achieved impressive successes in a broad range of areas, such as coordination of robotic swarms (Preiss et al., 2017), self-driving vehicles (DeWeese & Qu, 2024), real-time bidding (Jin et al., 2018), ride-sharing (Li et al., 2019), and stochastic games (Jin et al., 2020). Despite growing interest in multi-agent RL (MARL), extending RL to multi-agent settings poses significant computational challenges due to the curse of dimensionality (Sayin et al., 2021). Even if the individual agents' state or action spaces are small, the global state space or action space can take values from a set with size that is exponentially large as a function of the number of agents.