Unlike the standard Reinforcement Learning (RL) model, many real-world tasks are non-Markovian, whose rewards are predicated on state history rather than solely on the current state. Solving a non-Markovian task, frequently applied in practical applications such as autonomous driving, financial trading, and medical diagnosis, can be quite challenging. We propose a novel RL approach to achieve non-Markovian rewards expressed in temporal logic LTL$_f$ (Linear Temporal Logic over Finite Traces). To this end, an encoding of linear complexity from LTL$_f$ into MDPs (Markov Decision Processes) is introduced to take advantage of advanced RL algorithms. Then, a prioritized experience replay technique based on the automata structure (semantics equivalent to LTL$_f$ specification) is utilized to improve the training process. We empirically evaluate several benchmark problems augmented with non-Markovian tasks to demonstrate the feasibility and effectiveness of our approach.

The Traveling Salesman Problem (TSP) is a well-known combinatorial optimization problem with broad real-world applications. Recently, neural networks have gained popularity in this research area because they provide strong heuristic solutions to TSPs. Compared to autoregressive neural approaches, non-autoregressive (NAR) networks exploit the inference parallelism to elevate inference speed but suffer from comparatively low solution quality. In this paper, we propose a novel NAR model named NAR4TSP, which incorporates a specially designed architecture and an enhanced reinforcement learning strategy. To the best of our knowledge, NAR4TSP is the first TSP solver that successfully combines RL and NAR networks. The key lies in the incorporation of NAR network output decoding into the training process. NAR4TSP efficiently represents TSP encoded information as rewards and seamlessly integrates it into reinforcement learning strategies, while maintaining consistent TSP sequence constraints during both training and testing phases. Experimental results on both synthetic and real-world TSP instances demonstrate that NAR4TSP outperforms four state-of-the-art models in terms of solution quality, inference speed, and generalization to unseen scenarios.

Safety is one of the biggest concerns to applying reinforcement learning (RL) to the physical world. In its core part, it is challenging to ensure RL agents persistently satisfy a hard state constraint without white-box or black-box dynamics models. This paper presents an integrated model learning and safe control framework to safeguard any agent, where its dynamics are learned as Gaussian processes. The proposed theory provides (i) a novel method to construct an offline dataset for model learning that best achieves safety requirements; (ii) a parameterization rule for safety index to ensure the existence of safe control; (iii) a safety guarantee in terms of probabilistic forward invariance when the model is learned using the aforementioned dataset. Simulation results show that our framework guarantees almost zero safety violation on various continuous control tasks.

Envisioned application areas for reinforcement learning (RL) include autonomous driving, precision agriculture, and finance, which all require RL agents to make decisions in the real world. A significant challenge hindering the adoption of RL methods in these domains is the non-robustness of conventional algorithms. In this paper, we argue that a fundamental issue contributing to this lack of robustness lies in the focus on the expected value of the return as the sole "correct" optimization objective. The expected value is the average over the statistical ensemble of infinitely many trajectories. For non-ergodic returns, this average differs from the average over a single but infinitely long trajectory. Consequently, optimizing the expected value can lead to policies that yield exceptionally high returns with probability zero but almost surely result in catastrophic outcomes. This problem can be circumvented by transforming the time series of collected returns into one with ergodic increments. This transformation enables learning robust policies by optimizing the long-term return for individual agents rather than the average across infinitely many trajectories. We propose an algorithm for learning ergodicity transformations from data and demonstrate its effectiveness in an instructive, non-ergodic environment and on standard RL benchmarks.

The combination of deep reinforcement learning (DRL) with ensemble methods has been proved to be highly effective in addressing complex sequential decision-making problems. This success can be primarily attributed to the utilization of multiple models, which enhances both the robustness of the policy and the accuracy of value function estimation. However, there has been limited analysis of the empirical success of current ensemble RL methods thus far. Our new analysis reveals that the sample efficiency of previous ensemble DRL algorithms may be limited by sub-policies that are not as diverse as they could be. Motivated by these findings, our study introduces a new ensemble RL algorithm, termed \textbf{T}rajectories-awar\textbf{E} \textbf{E}nsemble exploratio\textbf{N} (TEEN). The primary goal of TEEN is to maximize the expected return while promoting more diverse trajectories. Through extensive experiments, we demonstrate that TEEN not only enhances the sample diversity of the ensemble policy compared to using sub-policies alone but also improves the performance over ensemble RL algorithms. On average, TEEN outperforms the baseline ensemble DRL algorithms by 41\% in performance on the tested representative environments.

Learning-based adaptive control methods hold the premise of enabling autonomous agents to reduce the effect of process variations with minimal human intervention. However, its application to autonomous underwater vehicles (AUVs) has so far been restricted due to 1) unknown dynamics under the form of sea current disturbance that we can not model properly nor measure due to limited sensor capability and 2) the nonlinearity of AUVs tasks where the controller response at some operating points must be overly conservative in order to satisfy the specification at other operating points. Deep Reinforcement Learning (DRL) can alleviates these limitations by training general-purpose neural network policies, but applications of DRL algorithms to AUVs have been restricted to simulated environments, due to their inherent high sample complexity and distribution shift problem. This paper presents a novel approach, merging the Maximum Entropy Deep Reinforcement Learning framework with a classic model-based control architecture, to formulate an adaptive controller. Within this framework, we introduce a Sim-to-Real transfer strategy comprising the following components: a bio-inspired experience replay mechanism, an enhanced domain randomisation technique, and an evaluation protocol executed on a physical platform. Our experimental assessments demonstrate that this method effectively learns proficient policies from suboptimal simulated models of the AUV, resulting in control performance 3 times higher when transferred to a real-world vehicle, compared to its model-based nonadaptive but optimal counterpart.

We propose BOSS, an approach that automatically learns to solve new long-horizon, complex, and meaningful tasks by growing a learned skill library with minimal supervision. Prior work in reinforcement learning require expert supervision, in the form of demonstrations or rich reward functions, to learn long-horizon tasks. Instead, our approach BOSS (BOotStrapping your own Skills) learns to accomplish new tasks by performing "skill bootstrapping," where an agent with a set of primitive skills interacts with the environment to practice new skills without receiving reward feedback for tasks outside of the initial skill set. This bootstrapping phase is guided by large language models (LLMs) that inform the agent of meaningful skills to chain together. Through this process, BOSS builds a wide range of complex and useful behaviors from a basic set of primitive skills. We demonstrate through experiments in realistic household environments that agents trained with our LLM-guided bootstrapping procedure outperform those trained with naive bootstrapping as well as prior unsupervised skill acquisition methods on zero-shot execution of unseen, long-horizon tasks in new environments. Website at clvrai.com/boss.

In this work, we present a scalable reinforcement learning method for training multi-task policies from large offline datasets that can leverage both human demonstrations and autonomously collected data. Our method uses a Transformer to provide a scalable representation for Q-functions trained via offline temporal difference backups. We therefore refer to the method as Q-Transformer. By discretizing each action dimension and representing the Q-value of each action dimension as separate tokens, we can apply effective high-capacity sequence modeling techniques for Q-learning. We present several design decisions that enable good performance with offline RL training, and show that Q-Transformer outperforms prior offline RL algorithms and imitation learning techniques on a large diverse real-world robotic manipulation task suite. The project's website and videos can be found at https://qtransformer.github.io

Although models trained on these datasets have shown strong performance in a wide variety of tasks such as classification, detection, and segmentation, they fundamentally are unable to generalize to an ever-evolving world due to constant out-of-distribution shifts of input data. Therefore, instead of training on fixed datasets, can we approach learning in a more human-centric and adaptive manner? In this paper, we introduce Action-Aware Embodied Learning for Perception (ALP), an embodied learning framework that incorporates action information into representation learning through a combination of optimizing a reinforcement learning policy and an inverse dynamics prediction objective. Our method actively explores in complex 3D environments to both learn generalizable task-agnostic visual representations as well as collect downstream training data. We show that ALP outperforms existing baselines in several downstream perception tasks. In addition, we show that by training on actively collected data more relevant to the environment and task, our method generalizes more robustly to downstream tasks compared to models pre-trained on fixed datasets such as ImageNet. A vast majority of current vision models are fueled by learning from a passively curated dataset. They achieve good performance on new inputs that are close to the training distribution but fail to generalize across changing conditions. This issue is mitigated partially by performing pre-training on a large dataset like ImageNet or COCO. However, these are still static snapshots of objects which do not allow for richer learning, and pre-trained models still fail to enable generalization to new environments and tasks that are semantically unrelated to those in the large-scale datasets. A natural approach to address these issues is to leverage active agents in an embodied environment. Intuitively, learning representations from exploration in the environment allows us to learn from large scale data that is relevant to the task at hand; active exploration can find maximally informative or diverse instances. A similar argument can be made for active data collection of labeled data for downstream fine-tuning.

One of the fundamental challenges for offline reinforcement learning (RL) is ensuring robustness to data distribution. Whether the data originates from a near-optimal policy or not, we anticipate that an algorithm should demonstrate its ability to learn an effective control policy that seamlessly aligns with the inherent distribution of offline data. Unfortunately, behavior regularization, a simple yet effective offline RL algorithm, tends to struggle in this regard. In this paper, we propose a new algorithm that substantially enhances behavior-regularization based on conservative policy iteration. Our key observation is that by iteratively refining the reference policy used for behavior regularization, conservative policy update guarantees gradually improvement, while also implicitly avoiding querying out-of-sample actions to prevent catastrophic learning failures. We prove that in the tabular setting this algorithm is capable of learning the optimal policy covered by the offline dataset, commonly referred to as the in-sample optimal policy. We then explore several implementation details of the algorithm when function approximations are applied. The resulting algorithm is easy to implement, requiring only a few lines of code modification to existing methods. Experimental results on the D4RL benchmark indicate that our method outperforms previous state-of-the-art baselines in most tasks, clearly demonstrate its superiority over behavior regularization.