Autonomous Vehicles (AVs), furnished with sensors capable of capturing essential vehicle dynamics such as speed, acceleration, and precise location, possess the capacity to execute intelligent maneuvers, including lane changes, in anticipation of approaching roadblocks. Nevertheless, the sheer volume of sensory data and the processing necessary to derive informed decisions can often overwhelm the vehicles, rendering them unable to handle the task independently. Consequently, a common approach in traffic scenarios involves transmitting the data to servers for processing, a practice that introduces challenges, particularly in situations demanding real-time processing. In response to this challenge, we present a novel DL-based semantic traffic control system that entrusts semantic encoding responsibilities to the vehicles themselves. This system processes driving decisions obtained from a Reinforcement Learning (RL) agent, streamlining the decision-making process. Specifically, our framework envisions scenarios where abrupt roadblocks materialize due to factors such as road maintenance, accidents, or vehicle repairs, necessitating vehicles to make determinations concerning lane-keeping or lane-changing actions to navigate past these obstacles. To formulate this scenario mathematically, we employ a Markov Decision Process (MDP) and harness the Deep Q Learning (DQN) algorithm to unearth viable solutions.

Mixtures of Experts (MoEs) have gained prominence in (self-)supervised learning due to their enhanced inference efficiency, adaptability to distributed training, and modularity. Previous research has illustrated that MoEs can significantly boost Deep Reinforcement Learning (DRL) performance by expanding the network's parameter count while reducing dormant neurons, thereby enhancing the model's learning capacity and ability to deal with non-stationarity. In this work, we shed more light on MoEs' ability to deal with non-stationarity and investigate MoEs in DRL settings with "amplified" non-stationarity via multi-task training, providing further evidence that MoEs improve learning capacity. In contrast to previous work, our multi-task results allow us to better understand the underlying causes for the beneficial effect of MoE in DRL training, the impact of the various MoE components, and insights into how best to incorporate them in actor-critic-based DRL networks. Finally, we also confirm results from previous work.

Reinforcement learning (RL) has proven to be well-performed and general-purpose in the inventory control (IC). However, further improvement of RL algorithms in the IC domain is impeded due to two limitations of online experience. First, online experience is expensive to acquire in real-world applications. With the low sample efficiency nature of RL algorithms, it would take extensive time to train the RL policy to convergence. Second, online experience may not reflect the true demand due to the lost sales phenomenon typical in IC, which makes the learning process more challenging. To address the above challenges, we propose a decision framework that combines reinforcement learning with feedback graph (RLFG) and intrinsically motivated exploration (IME) to boost sample efficiency. In particular, we first take advantage of the inherent properties of lost-sales IC problems and design the feedback graph (FG) specially for lost-sales IC problems to generate abundant side experiences aid RL updates. Then we conduct a rigorous theoretical analysis of how the designed FG reduces the sample complexity of RL methods. Based on the theoretical insights, we design an intrinsic reward to direct the RL agent to explore to the state-action space with more side experiences, further exploiting FG's power. Experimental results demonstrate that our method greatly improves the sample efficiency of applying RL in IC. Our code is available at https://anonymous.4open.science/r/RLIMFG4IC-811D/

Applying reinforcement learning (RL) to real-world problems is often made challenging by the inability to interact with the environment and the difficulty of designing reward functions. Offline RL addresses the first challenge by considering access to an offline dataset of environment interactions labeled by the reward function. In contrast, Preference-based RL does not assume access to the reward function and learns it from preferences, but typically requires an online interaction with the environment. We bridge the gap between these frameworks by exploring efficient methods for acquiring preference feedback in a fully offline setup. We propose Sim-OPRL, an offline preference-based reinforcement learning algorithm, which leverages a learned environment model to elicit preference feedback on simulated rollouts. Drawing on insights from both the offline RL and the preference-based RL literature, our algorithm employs a pessimistic approach for out-of-distribution data, and an optimistic approach for acquiring informative preferences about the optimal policy. We provide theoretical guarantees regarding the sample complexity of our approach, dependent on how well the offline data covers the optimal policy. Finally, we demonstrate the empirical performance of Sim-OPRL in different environments.

The constrained Markov decision process (CMDP) framework emerges as an important reinforcement learning approach for imposing safety or other critical objectives while maximizing cumulative reward. However, the current understanding of how to learn efficiently in a CMDP environment with a potentially infinite number of states remains under investigation, particularly when function approximation is applied to the value functions. In this paper, we address the learning problem given linear function approximation with $q_{\pi}$-realizability, where the value functions of all policies are linearly representable with a known feature map, a setting known to be more general and challenging than other linear settings. Utilizing a local-access model, we propose a novel primal-dual algorithm that, after $\tilde{O}(\text{poly}(d) \epsilon^{-3})$ queries, outputs with high probability a policy that strictly satisfies the constraints while nearly optimizing the value with respect to a reward function. Here, $d$ is the feature dimension and $\epsilon > 0$ is a given error. The algorithm relies on a carefully crafted off-policy evaluation procedure to evaluate the policy using historical data, which informs policy updates through policy gradients and conserves samples. To our knowledge, this is the first result achieving polynomial sample complexity for CMDP in the $q_{\pi}$-realizable setting.

Robustness remains a paramount concern in deep reinforcement learning (DRL), with randomized smoothing emerging as a key technique for enhancing this attribute. However, a notable gap exists in the performance of current smoothed DRL agents, often characterized by significantly low clean rewards and weak robustness. In response to this challenge, our study introduces innovative algorithms aimed at training effective smoothed robust DRL agents. We propose S-DQN and S-PPO, novel approaches that demonstrate remarkable improvements in clean rewards, empirical robustness, and robustness guarantee across standard RL benchmarks. Notably, our S-DQN and S-PPO agents not only significantly outperform existing smoothed agents by an average factor of $2.16\times$ under the strongest attack, but also surpass previous robustly-trained agents by an average factor of $2.13\times$. This represents a significant leap forward in the field. Furthermore, we introduce Smoothed Attack, which is $1.89\times$ more effective in decreasing the rewards of smoothed agents than existing adversarial attacks.

Pathology image are essential for accurately interpreting lesion cells in cytopathology screening, but acquiring high-resolution digital slides requires specialized equipment and long scanning times. Though super-resolution (SR) techniques can alleviate this problem, existing deep learning models recover pathology image in a black-box manner, which can lead to untruthful biological details and misdiagnosis. Additionally, current methods allocate the same computational resources to recover each pixel of pathology image, leading to the sub-optimal recovery issue due to the large variation of pathology image. In this paper, we propose the first hierarchical reinforcement learning framework named Spatial-Temporal hierARchical Reinforcement Learning (STAR-RL), mainly for addressing the aforementioned issues in pathology image super-resolution problem. We reformulate the SR problem as a Markov decision process of interpretable operations and adopt the hierarchical recovery mechanism in patch level, to avoid sub-optimal recovery. Specifically, the higher-level spatial manager is proposed to pick out the most corrupted patch for the lower-level patch worker. Moreover, the higher-level temporal manager is advanced to evaluate the selected patch and determine whether the optimization should be stopped earlier, thereby avoiding the over-processed problem. Under the guidance of spatial-temporal managers, the lower-level patch worker processes the selected patch with pixel-wise interpretable actions at each time step. Experimental results on medical images degraded by different kernels show the effectiveness of STAR-RL. Furthermore, STAR-RL validates the promotion in tumor diagnosis with a large margin and shows generalizability under various degradations. The source code is available at https://github.com/CUHK-AIM-Group/STAR-RL.

Can emergent language models faithfully model the intelligence of decision-making agents? Though modern language models exhibit already some reasoning ability, and theoretically can potentially express any probable distribution over tokens, it remains underexplored how the world knowledge these pretrained models have memorized can be utilized to comprehend an agent's behaviour in the physical world. This study empirically examines, for the first time, how well large language models (LLMs) can build a mental model of agents, termed agent mental modelling, by reasoning about an agent's behaviour and its effect on states from agent interaction history. This research may unveil the potential of leveraging LLMs for elucidating RL agent behaviour, addressing a key challenge in eXplainable reinforcement learning (XRL). To this end, we propose specific evaluation metrics and test them on selected RL task datasets of varying complexity, reporting findings on agent mental model establishment. Our results disclose that LLMs are not yet capable of fully mental modelling agents through inference alone without further innovations. This work thus provides new insights into the capabilities and limitations of modern LLMs.

Connected automated driving promises a significant improvement of traffic efficiency and safety on highways and in urban areas. Cooperative maneuver planning may facilitate active guidance of connected automated vehicles at intersections. Research in automatic intersection management put forth a large body of works that mostly employ rule-based or optimization-based approaches primarily in fully automated simulated environments. In this work, we compare two cooperative planning approaches for unsignalized intersections that are capable of handling mixed traffic, i.e., the road being shared by automated vehicles and regular vehicles driven by humans. The first approach is a cooperative planner that selects the most efficient out of multiple possible maneuvers based on a scene prediction trained on real driving data. The second cooperative planning approach is based on graph-based reinforcement learning, which conquers the lack of ground truth data for cooperative maneuvers. We thoroughly evaluate both cooperative planners in a realistic high-fidelity simulation with fully automated traffic and mixed traffic. The simulative experiments show that cooperative maneuver planning leads to less delay due to interaction and a reduced number of stops. Furthermore, we present results from real-world experiments with three prototype automated vehicles at a T-junction in public traffic, in which both planning modules demonstrate their ability to perform efficient cooperative maneuvers.

Recovering reward function from expert demonstrations is a fundamental problem in reinforcement learning. The recovered reward function captures the motivation of the expert. Agents can imitate experts by following these reward functions in their environment, which is known as apprentice learning. However, the agents may face environments different from the demonstrations, and therefore, desire transferable reward functions. Classical reward learning methods such as inverse reinforcement learning (IRL) or, equivalently, adversarial imitation learning (AIL), recover reward functions coupled with training dynamics, which are hard to be transferable. Previous dynamics-agnostic reward learning methods rely on assumptions such as that the reward function has to be state-only, restricting their applicability. In this work, we present a dynamics-agnostic discriminator-ensemble reward learning method (DARL) within the AIL framework, capable of learning both state-action and state-only reward functions. DARL achieves this by decoupling the reward function from training dynamics, employing a dynamics-agnostic discriminator on a latent space derived from the original state-action space. This latent space is optimized to minimize information on the dynamics. We moreover discover the policy-dependency issue of the AIL framework that reduces the transferability. DARL represents the reward function as an ensemble of discriminators during training to eliminate policy dependencies. Empirical studies on MuJoCo tasks with changed dynamics show that DARL better recovers the reward function and results in better imitation performance in transferred environments, handling both state-only and state-action reward scenarios.