We propose Deterministic Sequencing of Exploration and Exploitation (DSEE) algorithm with interleaving exploration and exploitation epochs for model-based RL problems that aim to simultaneously learn the system model, i.e., a Markov decision process (MDP), and the associated optimal policy. During exploration, DSEE explores the environment and updates the estimates for expected reward and transition probabilities. During exploitation, the latest estimates of the expected reward and transition probabilities are used to obtain a robust policy with high probability. We design the lengths of the exploration and exploitation epochs such that the cumulative regret grows as a sub-linear function of time.

When studying the potential advantages of quantum computing in machine learning, a natural question that arises is whether quantum algorithms that exploit quantum access to data can speed up learning. In the context of supervised learning, this led to the development of algorithms based on quantum RAMs, which can achieve high-degree polynomial improvements over their classical analogs [1]. In reinforcement learning, where we consider learning agents interacting with task environments, the question becomes: can quantum interactions with an environment, and in particular the ability to explore several trajectories in superposition, be beneficial for a learning agent. In recent years, several works have approached this question from a variety of angles [2]: based on Grover's algorithm [3], some works have for instance shown that searching for an optimal sequence of actions in an environment can be done using quadratically fewer interactions given the appropriate oracular access to the environment [4, 5, 6]. Other works have considered the more general problem of finding the optimal policy in a Markov Decision Process (MDP), and have found that up to quadratic speed-ups in the number of interactions are also possible, again given the proper oracular access [7, 8, 9, 10]. Finally, tailored MDP environments (based, e.g., on Simon's problem) have also been introduced, which allow for exponential quantum speed-ups in learning times compared to the best classical agents [11]. Yet, all the quantum algorithms that have been proposed in this quantum-accessible setting remain inefficient in the most well-publicised use cases of reinforcement learning, such as Go [12], city navigation [13], and computer games [14]: environments with large state-action spaces. Indeed, aside from the task-specific algorithms of Ref. [11], the proposed algorithms scale at best as the square root of the size of the state-action space, which is intractable in most modern-day applications that deal for instance with image-based inputs. In the classical literature, modern approaches to reinforcement learning in large spaces commonly replace the explicit storage of a policy (and/or a value function) in a table of values by a parametrized model (e.g., a

We consider the problem of two active particles in 2D complex flows with the multi-objective goals of minimizing both the dispersion rate and the energy consumption of the pair. We approach the problem by means of Multi Objective Reinforcement Learning (MORL), combining scalarization techniques together with a Q-learning algorithm, for Lagrangian drifters that have variable swimming velocity. We show that MORL is able to find a set of trade-off solutions forming an optimal Pareto frontier. As a benchmark, we show that a set of heuristic strategies are dominated by the MORL solutions. We consider the situation in which the agents cannot update their control variables continuously, but only after a discrete (decision) time, $\tau$. We show that there is a range of decision times, in between the Lyapunov time and the continuous updating limit, where Reinforcement Learning finds strategies that significantly improve over heuristics. In particular, we discuss how large decision times require enhanced knowledge of the flow, whereas for smaller $\tau$ all a priori heuristic strategies become Pareto optimal.

Complex and contact-rich robotic manipulation tasks, particularly those that involve multi-fingered hands and underactuated object manipulation, present a significant challenge to any control method. Methods based on reinforcement learning offer an appealing choice for such settings, as they can enable robots to learn to delicately balance contact forces and dexterously reposition objects without strong modeling assumptions. However, running reinforcement learning on real-world dexterous manipulation systems often requires significant manual engineering. This negates the benefits of autonomous data collection and ease of use that reinforcement learning should in principle provide. In this paper, we describe a system for vision-based dexterous manipulation that provides a "programming-free" approach for users to define new tasks and enable robots with complex multi-fingered hands to learn to perform them through interaction. The core principle underlying our system is that, in a vision-based setting, users should be able to provide high-level intermediate supervision that circumvents challenges in teleoperation or kinesthetic teaching which allow a robot to not only learn a task efficiently but also to autonomously practice. Our system includes a framework for users to define a final task and intermediate sub-tasks with image examples, a reinforcement learning procedure that learns the task autonomously without interventions, and experimental results with a four-finger robotic hand learning multi-stage object manipulation tasks directly in the real world, without simulation, manual modeling, or reward engineering.

Fictitious play is an algorithm for computing Nash equilibria of matrix games. Recently, machine learning variants of fictitious play have been successfully applied to complicated real-world games. This paper presents a simple modification of fictitious play which is a strict improvement over the original: it has the same theoretical worst-case convergence rate, is equally applicable in a machine learning context, and enjoys superior empirical performance. We conduct an extensive comparison of our algorithm with fictitious play, proving an optimal convergence rate for certain classes of games, demonstrating superior performance numerically across a variety of games, and concluding with experiments that extend these algorithms to the setting of deep multiagent reinforcement learning.

Abstractive dialogue summarization has received increasing attention recently. Despite the fact that most of the current dialogue summarization systems are trained to maximize the likelihood of human-written summaries and have achieved significant results, there is still a huge gap in generating high-quality summaries as determined by humans, such as coherence and faithfulness, partly due to the misalignment in maximizing a single human-written summary. To this end, we propose to incorporate different levels of human feedback into the training process. This will enable us to guide the models to capture the behaviors humans care about for summaries. Specifically, we ask humans to highlight the salient information to be included in summaries to provide the local feedback , and to make overall comparisons among summaries in terms of coherence, accuracy, coverage, concise and overall quality, as the global feedback. We then combine both local and global feedback to fine-tune the dialog summarization policy with Reinforcement Learning. Experiments conducted on multiple datasets demonstrate the effectiveness and generalization of our methods over the state-of-the-art supervised baselines, especially in terms of human judgments.

Many real-world reinforcement learning tasks require control of complex dynamical systems that involve both costly data acquisition processes and large state spaces. In cases where the transition dynamics can be readily evaluated at specified states (e.g., via a simulator), agents can operate in what is often referred to as planning with a generative model. We propose the AE-LSVI algorithm for bestpolicy identification, a novel variant of the kernelized least-squares value iteration (LSVI) algorithm that combines optimism with pessimism for active exploration (AE). AE-LSVI provably identifies a near-optimal policy uniformly over an entire state space and achieves polynomial sample complexity guarantees that are independent of the number of states. When specialized to the recently introduced offline contextual Bayesian optimization setting, our algorithm achieves improved sample complexity bounds. Experimentally, we demonstrate that AE-LSVI outperforms other RL algorithms in a variety of environments when robustness to the initial state is required. Reinforcement learning (RL) algorithms are increasingly applied to complex domains such as robotics (Kober et al., 2013), magnetic tokamaks (Seo et al., 2021; Degrave et al., 2022), and molecular search (Simm et al., 2020a;b). A central challenge in such environments is that data acquisition is often a time-consuming and expensive process, or may be infeasible due to safety considerations.

Recently, extensive studies on photonic reinforcement learning to accelerate the process of calculation by exploiting the physical nature of light have been conducted. Previous studies utilized quantum interference of photons to achieve collective decision-making without choice conflicts when solving the competitive multi-armed bandit problem, a fundamental example of reinforcement learning. However, the bandit problem deals with a static environment where the agent's action does not influence the reward probabilities. This study aims to extend the conventional approach to a more general multi-agent reinforcement learning targeting the grid world problem. Unlike the conventional approach, the proposed scheme deals with a dynamic environment where the reward changes because of agents' actions. A successful photonic reinforcement learning scheme requires both a photonic system that contributes to the quality of learning and a suitable algorithm. This study proposes a novel learning algorithm, discontinuous bandit Q-learning, in view of a potential photonic implementation. Here, state-action pairs in the environment are regarded as slot machines in the context of the bandit problem and an updated amount of Q-value is regarded as the reward of the bandit problem. We perform numerical simulations to validate the effectiveness of the bandit algorithm. In addition, we propose a multi-agent architecture in which agents are indirectly connected through quantum interference of light and quantum principles ensure the conflict-free property of state-action pair selections among agents. We demonstrate that multi-agent reinforcement learning can be accelerated owing to conflict avoidance among multiple agents.

The Metaverse can be considered the extension of the present-day web, which integrates the physical and virtual worlds, delivering hyper-realistic user experiences. The inception of the Metaverse brings forth many ecosystem services such as content creation, social entertainment, in-world value transfer, intelligent traffic, healthcare. These services are compute-intensive and require computation offloading onto a Metaverse edge computing server (MECS). Existing Metaverse edge computing approaches do not efficiently and effectively handle resource allocation to ensure a fluid, seamless and hyper-realistic Metaverse experience required for Metaverse ecosystem services. Therefore, we introduce a new Metaverse-compatible, Unified, User and Task (UUT) centered artificial intelligence (AI)- based mobile edge computing (MEC) paradigm, which serves as a concept upon which future AI control algorithms could be built to develop a more user and task-focused MEC.

Offline policy evaluation is a fundamental statistical problem in reinforcement learning that involves estimating the value function of some decision-making policy given data collected by a potentially different policy. In order to tackle problems with complex, high-dimensional observations, there has been significant interest from theoreticians and practitioners alike in understanding the possibility of function approximation in reinforcement learning. Despite significant study, a sharp characterization of when we might expect offline policy evaluation to be tractable, even in the simplest setting of linear function approximation, has so far remained elusive, with a surprising number of strong negative results recently appearing in the literature. In this work, we identify simple control-theoretic and linear-algebraic conditions that are necessary and sufficient for classical methods, in particular Fitted Q-iteration (FQI) and least squares temporal difference learning (LSTD), to succeed at offline policy evaluation. Using this characterization, we establish a precise hierarchy of regimes under which these estimators succeed. We prove that LSTD works under strictly weaker conditions than FQI. Furthermore, we establish that if a problem is not solvable via LSTD, then it cannot be solved by a broad class of linear estimators, even in the limit of infinite data. Taken together, our results provide a complete picture of the behavior of linear estimators for offline policy evaluation, unify previously disparate analyses of canonical algorithms, and provide significantly sharper notions of the underlying statistical complexity of offline policy evaluation.