Reinforcement learning (RL) agents are vulnerable to adversarial disturbances, which can deteriorate task performance or compromise safety specifications. Existing methods either address safety requirements under the assumption of no adversary (e.g., safe RL) or only focus on robustness against performance adversaries (e.g., robust RL). Learning one policy that is both safe and robust remains a challenging open problem. The difficulty is how to tackle two intertwined aspects in the worst cases: feasibility and optimality. Optimality is only valid inside a feasible region, while identification of maximal feasible region must rely on learning the optimal policy. To address this issue, we propose a systematic framework to unify safe RL and robust RL, including problem formulation, iteration scheme, convergence analysis and practical algorithm design. This unification is built upon constrained two-player zero-sum Markov games. A dual policy iteration scheme is proposed, which simultaneously optimizes a task policy and a safety policy. The convergence of this iteration scheme is proved. Furthermore, we design a deep RL algorithm for practical implementation, called dually robust actor-critic (DRAC). The evaluations with safety-critical benchmarks demonstrate that DRAC achieves high performance and persistent safety under all scenarios (no adversary, safety adversary, performance adversary), outperforming all baselines significantly.

Reinforcement learning (RL) is a popular technique that allows an agent to learn by trial and error while interacting with a dynamic environment. The traditional Reinforcement Learning (RL) approach has been successful in learning and predicting Euclidean robotic manipulation skills such as positions, velocities, and forces. However, in robotics, it is common to encounter non-Euclidean data such as orientation or stiffness, and failing to account for their geometric nature can negatively impact learning accuracy and performance. In this paper, to address this challenge, we propose a novel framework for RL that leverages Riemannian geometry, which we call Geometric Reinforcement Learning (G-RL), to enable agents to learn robotic manipulation skills with non-Euclidean data. Specifically, G-RL utilizes the tangent space in two ways: a tangent space for parameterization and a local tangent space for mapping to a nonEuclidean manifold. The policy is learned in the parameterization tangent space, which remains constant throughout the training. The policy is then transferred to the local tangent space via parallel transport and projected onto the non-Euclidean manifold. The local tangent space changes over time to remain within the neighborhood of the current manifold point, reducing the approximation error. Therefore, by introducing a geometrically grounded pre- and post-processing step into the traditional RL pipeline, our G-RL framework enables several model-free algorithms designed for Euclidean space to learn from non-Euclidean data without modifications. Experimental results, obtained both in simulation and on a real robot, support our hypothesis that G-RL is more accurate and converges to a better solution than approximating non-Euclidean data.

This report presents a comprehensive view of our vision on the development path of the human-machine symbiotic art creation. We propose a classification of the creative system with a hierarchy of 5 classes, showing the pathway of creativity evolving from a mimic-human artist (Turing Artists) to a Machine artist in its own right. We begin with an overview of the limitations of the Turing Artists then focus on the top two-level systems, Machine Artists, emphasizing machine-human communication in art creation. In art creation, it is necessary for machines to understand humans' mental states, including desires, appreciation, and emotions, humans also need to understand machines' creative capabilities and limitations. The rapid development of immersive environment and further evolution into the new concept of metaverse enable symbiotic art creation through unprecedented flexibility of bi-directional communication between artists and art manifestation environments. By examining the latest sensor and XR technologies, we illustrate the novel way for art data collection to constitute the base of a new form of human-machine bidirectional communication and understanding in art creation. Based on such communication and understanding mechanisms, we propose a novel framework for building future Machine artists, which comes with the philosophy that a human-compatible AI system should be based on the "human-in-the-loop" principle rather than the traditional "end-to-end" dogma. By proposing a new form of inverse reinforcement learning model, we outline the platform design of machine artists, demonstrate its functions and showcase some examples of technologies we have developed. We also provide a systematic exposition of the ecosystem for AI-based symbiotic art form and community with an economic model built on NFT technology. Ethical issues for the development of machine artists are also discussed.

We consider a decentralized formulation of the active hypothesis testing (AHT) problem, where multiple agents gather noisy observations from the environment with the purpose of identifying the correct hypothesis. At each time step, agents have the option to select a sampling action. These different actions result in observations drawn from various distributions, each associated with a specific hypothesis. The agents collaborate to accomplish the task, where message exchanges between agents are allowed over a rate-limited communications channel. The objective is to devise a multi-agent policy that minimizes the Bayes risk. This risk comprises both the cost of sampling and the joint terminal cost incurred by the agents upon making a hypothesis declaration. Deriving optimal structured policies for AHT problems is generally mathematically intractable, even in the context of a single agent. As a result, recent efforts have turned to deep learning methodologies to address these problems, which have exhibited significant success in single-agent learning scenarios. In this paper, we tackle the multi-agent AHT formulation by introducing a novel algorithm rooted in the framework of deep multi-agent reinforcement learning. This algorithm, named Multi-Agent Reinforcement Learning for AHT (MARLA), operates at each time step by having each agent map its state to an action (sampling rule or stopping rule) using a trained deep neural network with the goal of minimizing the Bayes risk. We present a comprehensive set of experimental results that effectively showcase the agents' ability to learn collaborative strategies and enhance performance using MARLA. Furthermore, we demonstrate the superiority of MARLA over single-agent learning approaches. Finally, we provide an open-source implementation of the MARLA framework, for the benefit of researchers and developers in related domains.

In many real-world scenarios involving high-stakes and safety implications, a human decision-maker (HDM) may receive recommendations from an artificial intelligence while holding the ultimate responsibility of making decisions. In this letter, we develop an "adherence-aware Q-learning" algorithm to address this problem. The algorithm learns the "adherence level" that captures the frequency with which an HDM follows the recommended actions and derives the best recommendation policy in real time. We prove the convergence of the proposed Q-learning algorithm to the optimal value and evaluate its performance across various scenarios.

Open-ended learning benefits immensely from the use of symbolic methods for goal representation as they offer ways to structure knowledge for efficient and transferable learning. However, the existing Hierarchical Reinforcement Learning (HRL) approaches relying on symbolic reasoning are often limited as they require a manual goal representation. The challenge in autonomously discovering a symbolic goal representation is that it must preserve critical information, such as the environment dynamics. In this work, we propose a developmental mechanism for subgoal discovery via an emergent representation that abstracts (i.e., groups together) sets of environment states that have similar roles in the task. We create a HRL algorithm that gradually learns this representation along with the policies and evaluate it on navigation tasks to show the learned representation is interpretable and results in data efficiency.

In this work, we focus on a robotic unloading problem from visual observations, where robots are required to autonomously unload stacks of parcels using RGB-D images as their primary input source. While supervised and imitation learning have accomplished good results in these types of tasks, they heavily rely on labeled data, which are challenging to obtain in realistic scenarios. Our study aims to develop a sample efficient controller framework that can learn unloading tasks without the need for labeled data during the learning process. To tackle this challenge, we propose a hierarchical controller structure that combines a high-level decision-making module with classical motion control. The high-level module is trained using Deep Reinforcement Learning (DRL), wherein we incorporate a safety bias mechanism and design a reward function tailored to this task. Our experiments demonstrate that both these elements play a crucial role in achieving improved learning performance. Furthermore, to ensure reproducibility and establish a benchmark for future research, we provide free access to our code and simulation.

Offline reinforcement learning (RL) holds promise as a means to learn high-reward policies from a static dataset, without the need for further environment interactions. However, a key challenge in offline RL lies in effectively stitching portions of suboptimal trajectories from the static dataset while avoiding extrapolation errors arising due to a lack of support in the dataset. Existing approaches use conservative methods that are tricky to tune and struggle with multi-modal data (as we show) or rely on noisy Monte Carlo return-to-go samples for reward conditioning. In this work, we propose a novel approach that leverages the expressiveness of latent diffusion to model in-support trajectory sequences as compressed latent skills. This facilitates learning a Q-function while avoiding extrapolation error via batch-constraining. The latent space is also expressive and gracefully copes with multi-modal data. We show that the learned temporally-abstract latent space encodes richer task-specific information for offline RL tasks as compared to raw state-actions. This improves credit assignment and facilitates faster reward propagation during Q-learning. Our method demonstrates state-of-the-art performance on the D4RL benchmarks, particularly excelling in long-horizon, sparse-reward tasks.

In the dynamic and uncertain environments where reinforcement learning (RL) operates, risk management becomes a crucial factor in ensuring reliable decision-making. Traditional RL approaches, while effective in reward optimization, often overlook the landscape of potential risks. In response, this paper pioneers the integration of Optimal Transport (OT) theory with RL to create a risk-aware framework. Our approach modifies the objective function, ensuring that the resulting policy not only maximizes expected rewards but also respects risk constraints dictated by OT distances between state visitation distributions and the desired risk profiles. By leveraging the mathematical precision of OT, we offer a formulation that elevates risk considerations alongside conventional RL objectives. Our contributions are substantiated with a series of theorems, mapping the relationships between risk distributions, optimal value functions, and policy behaviors. Through the lens of OT, this work illuminates a promising direction for RL, ensuring a balanced fusion of reward pursuit and risk awareness.

Deep Reinforcement Learning (DRL) has achieved remarkable success in sequential decision-making problems. However, existing DRL agents make decisions in an opaque fashion, hindering the user from establishing trust and scrutinizing weaknesses of the agents. While recent research has developed Interpretable Policy Extraction (IPE) methods for explaining how an agent takes actions, their explanations are often inconsistent with the agent's behavior and thus, frequently fail to explain. To tackle this issue, we propose a novel method, Fidelity-Induced Policy Extraction (FIPE). Specifically, we start by analyzing the optimization mechanism of existing IPE methods, elaborating on the issue of ignoring consistency while increasing cumulative rewards. We then design a fidelity-induced mechanism by integrate a fidelity measurement into the reinforcement learning feedback. We conduct experiments in the complex control environment of StarCraft II, an arena typically avoided by current IPE methods. The experiment results demonstrate that FIPE outperforms the baselines in terms of interaction performance and consistency, meanwhile easy to understand.