In various decision-making problems, estimating the value, the expected reward of a policy is a crucial question that needs to be addressed. Online evaluation requiring a comprehensive evaluation of policy value can be expensive and may not be applicable to multiple target policies. Alternatively, off-policy evaluation (OPE) refers to a technique that estimates the value of a target policy by utilizing log data generated from a different logging policy. This approach has attracted considerable interest in the domains of contextual bandits (CB) [Dudík et al., 2011, Swaminathan et al., 2017] and reinforcement learning (RL) [Precup, 2000, Mahmood et al., 2014, Jiang and Li, 2016]. Several off-policy evaluation algorithms [Dudík et al., 2011, Thomas and Brunskill, 2016, Wang et al., 2017, Farajtabar et al., 2018, Su et al., 2020] currently in use rely on having complete knowledge of the logging policy in order to utilize inverse probability weighting (IPW).

Let $p$ denote a generative language model. Let $r$ denote a reward model that returns a scalar that captures the degree at which a draw from $p$ is preferred. The goal of language model alignment is to alter $p$ to a new distribution $\phi$ that results in a higher expected reward while keeping $\phi$ close to $p.$ A popular alignment method is the KL-constrained reinforcement learning (RL), which chooses a distribution $\phi_\Delta$ that maximizes $E_{\phi_{\Delta}} r(y)$ subject to a relative entropy constraint $KL(\phi_\Delta || p) \leq \Delta.$ Another simple alignment method is best-of-$N$, where $N$ samples are drawn from $p$ and one with highest reward is selected. In this paper, we offer a closed-form characterization of the optimal KL-constrained RL solution. We demonstrate that any alignment method that achieves a comparable trade-off between KL divergence and reward must approximate the optimal KL-constrained RL solution in terms of relative entropy. To further analyze the properties of alignment methods, we introduce two simplifying assumptions: we let the language model be memoryless, and the reward model be linear. Although these assumptions may not reflect complex real-world scenarios, they enable a precise characterization of the asymptotic behavior of both the best-of-$N$ alignment, and the KL-constrained RL method, in terms of information-theoretic quantities. We prove that the reward of the optimal KL-constrained RL solution satisfies a large deviation principle, and we fully characterize its rate function. We also show that the rate of growth of the scaled cumulants of the reward is characterized by a proper Renyi cross entropy. Finally, we show that best-of-$N$ is asymptotically equivalent to KL-constrained RL solution by proving that their expected rewards are asymptotically equal, and concluding that the two distributions must be close in KL divergence.

Effective operation and seamless cooperation of robotic systems are a fundamental component of next-generation technologies and applications. In contexts such as disaster response, swarm operations require coordinated behavior and mobility control to be handled in a distributed manner, with the quality of the agents' actions heavily relying on the communication between them and the underlying network. In this paper, we formulate the problem of dynamic network bridging in a novel Decentralized Partially Observable Markov Decision Process (Dec-POMDP), where a swarm of agents cooperates to form a link between two distant moving targets. Furthermore, we propose a Multi-Agent Reinforcement Learning (MARL) approach for the problem based on Graph Convolutional Reinforcement Learning (DGN) which naturally applies to the networked, distributed nature of the task. The proposed method is evaluated in a simulated environment and compared to a centralized heuristic baseline showing promising results. Moreover, a further step in the direction of sim-to-real transfer is presented, by additionally evaluating the proposed approach in a near Live Virtual Constructive (LVC) UAV framework.

Low earth orbit (LEO) satellite systems play an important role in next generation communication networks due to their ability to provide extensive global coverage with guaranteed communications in remote areas and isolated areas where base stations cannot be cost-efficiently deployed. With the pervasive adoption of LEO satellite systems, especially in the LEO Internet-of-Things (IoT) scenarios, their spectrum resource management requirements have become more complex as a result of massive service requests and high bandwidth demand from terrestrial terminals. For instance, when leasing the spectrum to terrestrial users and controlling the uplink transmit power, satellites collect user data for machine learning purposes, which usually are sensitive information such as location, budget and quality of service (QoS) requirement. To facilitate model training in LEO IoT while preserving the privacy of data, blockchain-driven federated learning (FL) is widely used by leveraging on a fully decentralized architecture. In this paper, we propose a hybrid spectrum pricing and power control framework for LEO IoT by combining blockchain technology and FL. We first design a local deep reinforcement learning algorithm for LEO satellite systems to learn a revenue-maximizing pricing and power control scheme. Then the agents collaborate to form a FL system. We also propose a reputation-based blockchain which is used in the global model aggregation phase of FL. Based on the reputation mechanism, a node is selected for each global training round to perform model aggregation and block generation, which can further enhance the decentralization of the network and guarantee the trust. Simulation tests are conducted to evaluate the performances of the proposed scheme. Our results show the efficiency of finding the maximum revenue scheme for LEO satellite systems while preserving the privacy of each agent.

Addressing complex cooperative tasks in safety-critical environments poses significant challenges for Multi-Agent Systems, especially under conditions of partial observability. This work introduces a hybrid approach that integrates Multi-Agent Reinforcement Learning with control-theoretic methods to ensure safe and efficient distributed strategies. Our contributions include a novel setpoint update algorithm that dynamically adjusts agents' positions to preserve safety conditions without compromising the mission's objectives. Through experimental validation, we demonstrate significant advantages over conventional MARL strategies, achieving comparable task performance with zero safety violations. Our findings indicate that integrating safe control with learning approaches not only enhances safety compliance but also achieves good performance in mission objectives.

Reinforcement Learning (RL) has progressed from simple control tasks to complex real-world challenges with large state spaces. While RL excels in these tasks, training time remains a limitation. Reward shaping is a popular solution, but existing methods often rely on value functions, which face scalability issues. This paper presents a novel safety-oriented reward-shaping framework inspired by barrier functions, offering simplicity and ease of implementation across various environments and tasks. To evaluate the effectiveness of the proposed reward formulations, we conduct simulation experiments on CartPole, Ant, and Humanoid environments, along with real-world deployment on the Unitree Go1 quadruped robot. Our results demonstrate that our method leads to 1.4-2.8 times faster convergence and as low as 50-60% actuation effort compared to the vanilla reward. In a sim-to-real experiment with the Go1 robot, we demonstrated better control and dynamics of the bot with our reward framework.

Humanoid Reaction Synthesis is pivotal for creating highly interactive and empathetic robots that can seamlessly integrate into human environments, enhancing the way we live, work, and communicate. However, it is difficult to learn the diverse interaction patterns of multiple humans and generate physically plausible reactions. The kinematics-based approaches face challenges, including issues like floating feet, sliding, penetration, and other problems that defy physical plausibility. The existing physics-based method often relies on kinematics-based methods to generate reference states, which struggle with the challenges posed by kinematic noise during action execution. Constrained by their reliance on diffusion models, these methods are unable to achieve real-time inference. In this work, we propose a Forward Dynamics Guided 4D Imitation method to generate physically plausible human-like reactions. The learned policy is capable of generating physically plausible and human-like reactions in real-time, significantly improving the speed(x33) and quality of reactions compared with the existing method. Our experiments on the InterHuman and Chi3D datasets, along with ablation studies, demonstrate the effectiveness of our approach.

Multi-robot collaborative navigation is an essential ability where teamwork and synchronization are keys. In complex and uncertain environments, adaptive formation is vital, as rigid formations prove to be inadequate. The ability of robots to dynamically adjust their formation enables navigation through unpredictable spaces, maintaining cohesion, and effectively responding to environmental challenges. In this paper, we introduce a novel approach that uses bi-level learning framework. Specifically, we use graph learning at a high level for group coordination and reinforcement learning for individual navigation. We innovate by integrating a spring-damper model within the reinforcement learning reward mechanism, addressing the rigidity of traditional formation control methods. During execution, our approach enables a team of robots to successfully navigate challenging environments, maintain a desired formation shape, and dynamically adjust their formation scale based on environmental information. We conduct extensive experiments to evaluate our approach across three distinct formation scenarios in multi-robot navigation: circle, line, and wedge. Experimental results show that our approach achieves promising results and scalability on multi-robot navigation with formation adaptation.

Autonomous assistance of people with motor impairments is one of the most promising applications of autonomous robotic systems. Recent studies have reported encouraging results using deep reinforcement learning (RL) in the healthcare domain. Previous studies showed that assistive tasks can be formulated as multi-agent RL, wherein there are two agents: a caregiver and a care-receiver. However, policies trained in multi-agent RL are often sensitive to the policies of other agents. In such a case, a trained caregiver's policy may not work for different care-receivers. To alleviate this issue, we propose a framework that learns a robust caregiver's policy by training it for diverse care-receiver responses. In our framework, diverse care-receiver responses are autonomously learned through trials and errors. In addition, to robustify the care-giver's policy, we propose a strategy for sampling a care-receiver's response in an adversarial manner during the training. We evaluated the proposed method using tasks in an Assistive Gym. We demonstrate that policies trained with a popular deep RL method are vulnerable to changes in policies of other agents and that the proposed framework improves the robustness against such changes.

Efficient exploration remains a challenging problem in reinforcement learning, especially for tasks where extrinsic rewards from environments are sparse or even totally disregarded. Significant advances based on intrinsic motivation show promising results in simple environments but often get stuck in environments with multimodal and stochastic dynamics. In this work, we propose a variational dynamic model based on the conditional variational inference to model the multimodality and stochasticity. We consider the environmental state-action transition as a conditional generative process by generating the next-state prediction under the condition of the current state, action, and latent variable, which provides a better understanding of the dynamics and leads a better performance in exploration. We derive an upper bound of the negative log-likelihood of the environmental transition and use such an upper bound as the intrinsic reward for exploration, which allows the agent to learn skills by self-supervised exploration without observing extrinsic rewards. We evaluate the proposed method on several image-based simulation tasks and a real robotic manipulating task. Our method outperforms several state-of-the-art environment model-based exploration approaches.