In this submission, a new bootstrapping optimization technique is proposed, based on the idea of adding the log-policy to the immediate reward. This is shown to bring strong empirical gains, and the theoretical analysis helps understand why. Although reviewers remained divided even after an active discussion period (7, 7, 5, 5), I believe this is a paper worth publishing at NeurIPS. Simple ideas bringing significant improvements, like this one, are typically those most impactful. I also appreciate the efforts made to better understand the theoretical properties of the proposed algorithm, beyond the basic intuition.

Additional Feedback: In the caption of figure 4 (b) I believe CNNs should be MLPs. In Figure captions it says that 25% 50% and 75% quantiles are shown but I only see one set of error bars. Line 142: Equation 9 should be Equation 8? Line 155: should this really be for all g given you are talking about a specific s \prime and a \prime? Is invertibility an assumption here? I can't immediately see why it should need to be so.

The paper proposes an approach for incorporating knowledge about symmetries or equivariances into neural network policies by providing a general purpose method for constructing network layers based on knowledge of the relevant transformations. The reviews are generally positive: Identifying effective ways of incorporating prior knowledge of this type into neural networks is an important research challenge that is of interest to the community. The proposed approach for constructing network layers seems novel, although there is some prior work that explores ways of exploiting such knowledge in particular application domains, or via alternative means such as data augmentation. An important caveat of the submission, remarked upon by all reviewers is the experimental evaluation. It is currently limited to simple scenarios with perfect symmetries which provide limited evidence of the utility of the approach in more complex / less idealized scenarios.

Additional Feedback: I really enjoyed reading the paper. The paper is clearly written and, in my opinion, proposes an elegant solution for an important problem in IRL---the fact that the same policy may arise from multiple rewards. There are, however, a couple of aspects that I think the paper could improve upon (several of which are just minor aspects). I number the different issues to facilitate author's responses. My main question is with respect to the use of the proposed projection with other forms of policy invariance.

Three out of four knowledgeable referees support acceptance for the contributions and I also recommend acceptance. I believe the concerns about theoretical aspects of R4 were addressed in the rebuttal. In the revised version of the paper, please present your additional experiments to address concerns of reviewers R3 and R4, your comments regarding runtime (R1) and your comments regarding the rho-projection (R1,R3,R4). Furthermore, R3 has some remaining theoretical concerns which were not clarified in the rebuttal - please elaborate on these in the revised paper.

The wind farm control problem is challenging, since conventional model-based control strategies require tractable models of complex aerodynamical interactions between the turbines and suffer from the curse of dimension when the number of turbines increases. Recently, model-free and multi-agent reinforcement learning approaches have been used to address this challenge. In this article, we introduce WFCRL (Wind Farm Control with Reinforcement Learning), the first open suite of multi-agent reinforcement learning environments for the wind farm control problem. WFCRL frames a cooperative Multi-Agent Reinforcement Learning (MARL) problem: each turbine is an agent and can learn to adjust its yaw, pitch or torque to maximize the common objective (e.g. the total power production of the farm). WFCRL also offers turbine load observations that will allow to optimize the farm performance while limiting turbine structural damages. Interfaces with two state-of-the-art farm simulators are implemented in WFCRL: a static simulator (FLORIS) and a dynamic simulator (FAST.Farm). For each simulator, $10$ wind layouts are provided, including $5$ real wind farms. Two state-of-the-art online MARL algorithms are implemented to illustrate the scaling challenges. As learning online on FAST.Farm is highly time-consuming, WFCRL offers the possibility of designing transfer learning strategies from FLORIS to FAST.Farm.

We present a Reinforcement Learning Platform for Adversarial Black-box untargeted and targeted attacks, RLAB, that allows users to select from various distortion filters to create adversarial examples. The platform uses a Reinforcement Learning agent to add minimum distortion to input images while still causing misclassification by the target model. The agent uses a novel dual-action method to explore the input image at each step to identify sensitive regions for adding distortions while removing noises that have less impact on the target model. This dual action leads to faster and more efficient convergence of the attack. The platform can also be used to measure the robustness of image classification models against specific distortion types. Also, retraining the model with adversarial samples significantly improved robustness when evaluated on benchmark datasets. The proposed platform outperforms state-of-the-art methods in terms of the average number of queries required to cause misclassification.

This paper introduces Localized Bipartite Match Graph Attention Q-Learning (BMG-Q), a novel Multi-Agent Reinforcement Learning (MARL) algorithm framework tailored for ride-pooling order dispatch. BMG-Q advances ride-pooling decision-making process with the localized bipartite match graph underlying the Markov Decision Process, enabling the development of novel Graph Attention Double Deep Q Network (GATDDQN) as the MARL backbone to capture the dynamic interactions among ride-pooling vehicles in fleet. Our approach enriches the state information for each agent with GATDDQN by leveraging a localized bipartite interdependence graph and enables a centralized global coordinator to optimize order matching and agent behavior using Integer Linear Programming (ILP). Enhanced by gradient clipping and localized graph sampling, our GATDDQN improves scalability and robustness. Furthermore, the inclusion of a posterior score function in the ILP captures the online exploration-exploitation trade-off and reduces the potential overestimation bias of agents, thereby elevating the quality of the derived solutions. Through extensive experiments and validation, BMG-Q has demonstrated superior performance in both training and operations for thousands of vehicle agents, outperforming benchmark reinforcement learning frameworks by around 10% in accumulative rewards and showing a significant reduction in overestimation bias by over 50%. Additionally, it maintains robustness amidst task variations and fleet size changes, establishing BMG-Q as an effective, scalable, and robust framework for advancing ride-pooling order dispatch operations.

We explore a capability of evolution strategies to train an agent with its policy based on a transformer architecture in a reinforcement learning setting. We performed experiments using OpenAI's highly parallelizable evolution strategy to train Decision Transformer in Humanoid locomotion environment and in the environment of Atari games, testing the ability of this black-box optimization technique to train even such relatively large and complicated models (compared to those previously tested in the literature). We also proposed a method to aid the training by first pretraining the model before using the OpenAI-ES to train it further, and tested its effectiveness. The examined evolution strategy proved to be, in general, capable of achieving strong results and managed to obtain high-performing agents. Therefore, the pretraining was shown to be unnecessary; yet still, it helped us observe and formulate several further insights.

This paper introduces a novel reinforcement learning (RL) framework, termed Reward-Guided Conservative Q-learning (RG-CQL), to enhance coordination between ride-pooling and public transit within a multimodal transportation network. We model each ride-pooling vehicle as an agent governed by a Markov Decision Process (MDP) and propose an offline training and online fine-tuning RL framework to learn the optimal operational decisions of the multimodal transportation systems, including rider-vehicle matching, selection of drop-off locations for passengers, and vehicle routing decisions, with improved data efficiency. During the offline training phase, we develop a Conservative Double Deep Q Network (CDDQN) as the action executor and a supervised learning-based reward estimator, termed the Guider Network, to extract valuable insights into action-reward relationships from data batches. In the online fine-tuning phase, the Guider Network serves as an exploration guide, aiding CDDQN in effectively and conservatively exploring unknown state-action pairs. The efficacy of our algorithm is demonstrated through a realistic case study using real-world data from Manhattan. We show that integrating ride-pooling with public transit outperforms two benchmark cases solo rides coordinated with transit and ride-pooling without transit coordination by 17% and 22% in the achieved system rewards, respectively. Furthermore, our innovative offline training and online fine-tuning framework offers a remarkable 81.3% improvement in data efficiency compared to traditional online RL methods with adequate exploration budgets, with a 4.3% increase in total rewards and a 5.6% reduction in overestimation errors. Experimental results further demonstrate that RG-CQL effectively addresses the challenges of transitioning from offline to online RL in large-scale ride-pooling systems integrated with transit.