Kernelized bandit (KB) problem [Srinivas et al., 2010], also called Gaussian process bandit or Bayesian optimization, is one of the important sequential decision-making problems where one seeks to minimize the regret under an unknown reward function via sequentially acquiring function evaluations. As the name suggests, in the KB problem, the underlying reward function is assumed to be an element of reproducing kernel Hilbert space (RKHS) induced by a known fixed kernel function. KB has been applied in many applications, such as materials discovery [Ueno et al., 2016], drug discovery [Korovina et al., 2020], and robotics [Berkenkamp et al., 2023]. In addition, the near-optimal KB algorithms, whose regret upper bound matches the regret lower bound derived in Scarlett et al. [2017], have been shown [Camilleri et al., 2021, Salgia et al., 2021, Li and Scarlett, 2022, Salgia et al., 2024]. Non-stationary KB [Bogunovic et al., 2016] considers the optimization under a non-stationary environment; that is, the reward function may change over time within some RKHS. This modification is crucial in many practical applications where an objective function varies over time, such as financial markets [Heaton and Lucas, 1999] and recommender systems [Hariri et al., 2015]. For example, Zhou and Shroff [2021], Deng et al. [2022] have proposed upper confidence bound (UCB)-based algorithms for the non-stationary KB problem and derived the upper bound of the cumulative regret. Recently, Hong et al. [2023] have proposed an optimization-based KB

Playing two-player games using reinforcement learning and self-play can be challenging due to the complexity of two-player environments and the possible instability in the training process. We propose that a reinforcement learning algorithm can train more efficiently and achieve improved performance in a two-player game if it leverages the knowledge from the single-player version of the same game. This study examines the proposed idea in ten different Atari 2600 environments using the Atari 2600 RAM as the input state. We discuss the advantages of using transfer learning from a single-player training process over training in a two-player setting from scratch, and demonstrate our results in a few measures such as training time and average total reward. We also discuss a method of calculating RAM complexity and its relationship to performance.

As artificial intelligence (AI) becomes deeply integrated into critical infrastructures and everyday life, ensuring its safe deployment is one of humanity's most urgent challenges. Current AI models prioritize task optimization over safety, leading to risks of unintended harm. These risks are difficult to address due to the competing interests of governments, businesses, and advocacy groups, all of which have different priorities in the AI race. Current alignment methods, such as reinforcement learning from human feedback (RLHF), focus on extrinsic behaviors without instilling a genuine understanding of human values. These models are vulnerable to manipulation and lack the social intelligence necessary to infer the mental states and intentions of others, raising concerns about their ability to safely and responsibly make important decisions in complex and novel situations. Furthermore, the divergence between extrinsic and intrinsic motivations in AI introduces the risk of deceptive or harmful behaviors, particularly as systems become more autonomous and intelligent. We propose a novel human-inspired approach which aims to address these various concerns and help align competing objectives.

Geomagnetic navigation has drawn increasing attention with its capacity in navigating through complex environments and its independence from external navigation services like global navigation satellite systems (GNSS). Existing studies on geomagnetic navigation, i.e., matching navigation and bionic navigation, rely on pre-stored map or extensive searches, leading to limited applicability or reduced navigation efficiency in unexplored areas. To address the issues with geomagnetic navigation in areas where GNSS is unavailable, this paper develops a deep reinforcement learning (DRL)-based mechanism, especially for long-distance geomagnetic navigation. The designed mechanism trains an agent to learn and gain the magnetoreception capacity for geomagnetic navigation, rather than using any pre-stored map or extensive and expensive searching approaches. Particularly, we integrate the geomagnetic gradient-based parallel approach into geomagnetic navigation. This integration mitigates the over-exploration of the learning agent by adjusting the geomagnetic gradient, such that the obtained gradient is aligned towards the destination. We explore the effectiveness of the proposed approach via detailed numerical simulations, where we implement twin delayed deep deterministic policy gradient (TD3) in realizing the proposed approach. The results demonstrate that our approach outperforms existing metaheuristic and bionic navigation methods in long-distance missions under diverse navigation conditions.

A novel class of advanced algorithms, termed Goal-Conditioned Weighted Supervised Learning (GCWSL), has recently emerged to tackle the challenges posed by sparse rewards in goal-conditioned reinforcement learning (RL). GCWSL consistently delivers strong performance across a diverse set of goal-reaching tasks due to its simplicity, effectiveness, and stability. However, GCWSL methods lack a crucial capability known as trajectory stitching, which is essential for learning optimal policies when faced with unseen skills during testing. This limitation becomes particularly pronounced when the replay buffer is predominantly filled with sub-optimal trajectories. In contrast, traditional TD-based RL methods, such as Q-learning, which utilize Dynamic Programming, do not face this issue but often experience instability due to the inherent difficulties in value function approximation. In this paper, we propose Q-learning Weighted Supervised Learning (Q-WSL), a novel framework designed to overcome the limitations of GCWSL by incorporating the strengths of Dynamic Programming found in Q-learning. Q-WSL leverages Dynamic Programming results to output the optimal action of (state, goal) pairs across different trajectories within the replay buffer. This approach synergizes the strengths of both Q-learning and GCWSL, effectively mitigating their respective weaknesses and enhancing overall performance. Empirical evaluations on challenging goal-reaching tasks demonstrate that Q-WSL surpasses other goal-conditioned approaches in terms of both performance and sample efficiency. Additionally, Q-WSL exhibits notable robustness in environments characterized by binary reward structures and environmental stochasticity.

Deep Reinforcement Learning (DRL) algorithms have achieved great success in solving many challenging tasks while their black-box nature hinders interpretability and real-world applicability, making it difficult for human experts to interpret and understand DRL policies. Existing works on interpretable reinforcement learning have shown promise in extracting decision tree (DT) based policies from DRL policies with most focus on the single-agent settings while prior attempts to introduce DT policies in multi-agent scenarios mainly focus on heuristic designs which do not provide any quantitative guarantees on the expected return. In this paper, we establish an upper bound on the return gap between the oracle expert policy and an optimal decision tree policy. This enables us to recast the DT extraction problem into a novel non-euclidean clustering problem over the local observation and action values space of each agent, with action values as cluster labels and the upper bound on the return gap as clustering loss. Both the algorithm and the upper bound are extended to multi-agent decentralized DT extractions by an iteratively-grow-DT procedure guided by an action-value function conditioned on the current DTs of other agents. Further, we propose the Return-Gap-Minimization Decision Tree (RGMDT) algorithm, which is a surprisingly simple design and is integrated with reinforcement learning through the utilization of a novel Regularized Information Maximization loss. Evaluations on tasks like D4RL show that RGMDT significantly outperforms heuristic DT-based baselines and can achieve nearly optimal returns under given DT complexity constraints (e.g., maximum number of DT nodes).

Since reinforcement learning algorithms are notoriously data-intensive, the task of sampling observations from the environment is usually split across multiple agents. However, transferring these observations from the agents to a central location can be prohibitively expensive in terms of communication cost, and it can also compromise the privacy of each agent's local behavior policy. Federated reinforcement learning is a framework in which $N$ agents collaboratively learn a global model, without sharing their individual data and policies. This global model is the unique fixed point of the average of $N$ local operators, corresponding to the $N$ agents. Each agent maintains a local copy of the global model and updates it using locally sampled data. In this paper, we show that by careful collaboration of the agents in solving this joint fixed point problem, we can find the global model $N$ times faster, also known as linear speedup. We first propose a general framework for federated stochastic approximation with Markovian noise and heterogeneity, showing linear speedup in convergence. We then apply this framework to federated reinforcement learning algorithms, examining the convergence of federated on-policy TD, off-policy TD, and $Q$-learning.

Autonomous navigation in unstructured outdoor environments is inherently challenging due to the presence of asymmetric traversal costs, such as varying energy expenditures for uphill versus downhill movement. Traditional reinforcement learning methods often assume symmetric costs, which can lead to suboptimal navigation paths and increased safety risks in real-world scenarios. In this paper, we introduce QuasiNav, a novel reinforcement learning framework that integrates quasimetric embeddings to explicitly model asymmetric costs and guide efficient, safe navigation. QuasiNav formulates the navigation problem as a constrained Markov decision process (CMDP) and employs quasimetric embeddings to capture directionally dependent costs, allowing for a more accurate representation of the terrain. This approach is combined with adaptive constraint tightening within a constrained policy optimization framework to dynamically enforce safety constraints during learning. We validate QuasiNav across three challenging navigation scenarios-undulating terrains, asymmetric hill traversal, and directionally dependent terrain traversal-demonstrating its effectiveness in both simulated and real-world environments. Experimental results show that QuasiNav significantly outperforms conventional methods, achieving higher success rates, improved energy efficiency, and better adherence to safety constraints.

Discovering meaningful insights from a large dataset, known as Exploratory Data Analysis (EDA), is a challenging task that requires thorough exploration and analysis of the data. Automated Data Exploration (ADE) systems use goal-oriented methods with Large Language Models and Reinforcement Learning towards full automation. However, these methods require human involvement to anticipate goals that may limit insight extraction, while fully automated systems demand significant computational resources and retraining for new datasets. We introduce QUIS, a fully automated EDA system that operates in two stages: insight generation (ISGen) driven by question generation (QUGen). The QUGen module generates questions in iterations, refining them from previous iterations to enhance coverage without human intervention or manually curated examples. The ISGen module analyzes data to produce multiple relevant insights in response to each question, requiring no prior training and enabling QUIS to adapt to new datasets.

Continual offline reinforcement learning (CORL) has shown impressive ability in diffusion-based lifelong learning systems by modeling the joint distributions of trajectories. However, most research only focuses on limited continual task settings where the tasks have the same observation and action space, which deviates from the realistic demands of training agents in various environments. In view of this, we propose Vector-Quantized Continual Diffuser, named VQ-CD, to break the barrier of different spaces between various tasks. Specifically, our method contains two complementary sections, where the quantization spaces alignment provides a unified basis for the selective weights activation. In the quantized spaces alignment, we leverage vector quantization to align the different state and action spaces of various tasks, facilitating continual training in the same space. Then, we propose to leverage a unified diffusion model attached by the inverse dynamic model to master all tasks by selectively activating different weights according to the task-related sparse masks. Finally, we conduct extensive experiments on 15 continual learning (CL) tasks, including conventional CL task settings (identical state and action spaces) and general CL task settings (various state and action spaces). Compared with 16 baselines, our method reaches the SOTA performance.