If an AI system makes decisions over time, how should we evaluate how aligned it is with a group of stakeholders (who may have conflicting values and preferences)? In this position paper, we advocate for consideration of temporal aspects including stakeholders' changing levels of satisfaction and their possibly temporally extended preferences. We suggest how a recent approach to evaluating fairness over time could be applied to a new form of pluralistic alignment: temporal pluralism, where the AI system reflects different stakeholders' values at different times.

In autonomous driving with image based state space, accurate prediction of future events and modeling diverse behavioral modes are essential for safety and effective decision-making. World model-based Reinforcement Learning (WMRL) approaches offers a promising solution by simulating future states from current state and actions. However, utility of world models is often limited by typical RL policies being limited to deterministic or single gaussian distribution. By failing to capture the full spectrum of possible actions, reduces their adaptability in complex, dynamic environments. In this work, we introduce Imagine-2-Drive, a framework that consists of two components, VISTAPlan, a high-fidelity world model for accurate future prediction and Diffusion Policy Actor (DPA), a diffusion based policy to model multi-modal behaviors for trajectory prediction. We use VISTAPlan to simulate and evaluate trajectories from DPA and use Denoising Diffusion Policy Optimization (DDPO) to train DPA to maximize the cumulative sum of rewards over the trajectories. We analyze the benefits of each component and the framework as a whole in CARLA with standard driving metrics. As a consequence of our twin novelties- VISTAPlan and DPA, we significantly outperform the state of the art (SOTA) world models on standard driving metrics by 15% and 20% on Route Completion and Success Rate respectively.

In this paper, we consider a wireless resource allocation problem in a cyber-physical system (CPS) where the control channel, carrying resource allocation commands, is subjected to denial-of-service (DoS) attacks. We propose a novel concept of collaborative distributed and centralized (CDC) resource allocation to effectively mitigate the impact of these attacks. To optimize the CDC resource allocation policy, we develop a new CDC-deep reinforcement learning (DRL) algorithm, whereas existing DRL frameworks only formulate either centralized or distributed decision-making problems. Simulation results demonstrate that the CDC-DRL algorithm significantly outperforms state-of-the-art DRL benchmarks, showcasing its ability to address resource allocation problems in large-scale CPSs under control channel attacks.

False data injection attacks (FDIAs) on smart inverters are a growing concern linked to increased renewable energy production. While data-based FDIA detection methods are also actively developed, we show that they remain vulnerable to impactful and stealthy adversarial examples that can be crafted using Reinforcement Learning (RL). We propose to include such adversarial examples in data-based detection training procedure via a continual adversarial RL (CARL) approach. This way, one can pinpoint the deficiencies of data-based detection, thereby offering explainability during their incremental improvement. We show that a continual learning implementation is subject to catastrophic forgetting, and additionally show that forgetting can be addressed by employing a joint training strategy on all generated FDIA scenarios.

Reinforcement learning (RL) is a sub-domain of machine learning, mainly concerned with solving sequential decision-making problems by a learning agent that interacts with the decision environment to improve its behavior through the reward it receives from the environment. This learning paradigm is, however, well-known for being time-consuming due to the necessity of collecting a large amount of data, making RL suffer from sample inefficiency and difficult generalization. Furthermore, the construction of an explicit reward function that accounts for the trade-off between multiple desiderata of a decision problem is often a laborious task. These challenges have been recently addressed utilizing transfer and inverse reinforcement learning (T-IRL). In this regard, this paper is devoted to a comprehensive review of realizing the sample efficiency and generalization of RL algorithms through T-IRL. Following a brief introduction to RL, the fundamental T-IRL methods are presented and the most recent advancements in each research field have been extensively reviewed. Our findings denote that a majority of recent research works have dealt with the aforementioned challenges by utilizing human-in-the-loop and sim-to-real strategies for the efficient transfer of knowledge from source domains to the target domain under the transfer learning scheme. Under the IRL structure, training schemes that require a low number of experience transitions and extension of such frameworks to multi-agent and multi-intention problems have been the priority of researchers in recent years.

Data-driven modeling for dynamic systems has gained widespread attention in recent years. Its inverse formulation, parameter estimation, aims to infer the inherent model parameters from observations. However, parameter degeneracy, where different combinations of parameters yield the same observable output, poses a critical barrier to accurately and uniquely identifying model parameters. In the context of WECC composite load model (CLM) in power systems, utility practitioners have observed that CLM parameters carefully selected for one fault event may not perform satisfactorily in another fault. Here, we innovate a joint conditional diffusion model-based inverse problem solver (JCDI), that incorporates a joint conditioning architecture with simultaneous inputs of multi-event observations to improve parameter generalizability. Simulation studies on the WECC CLM show that the proposed JCDI effectively reduces uncertainties of degenerate parameters, thus the parameter estimation error is decreased by 42.1% compared to a single-event learning scheme. This enables the model to achieve high accuracy in predicting power trajectories under different fault events, including electronic load tripping and motor stalling, outperforming standard deep reinforcement learning and supervised learning approaches. We anticipate this work will contribute to mitigating parameter degeneracy in system dynamics, providing a general parameter estimation framework across various scientific domains.

In 2020, we introduced a deep reinforcement learning method capable of generating superhuman chip layouts, which we then published in Nature and open-sourced on GitHub. AlphaChip has inspired an explosion of work on AI for chip design, and has been deployed in state-of-the-art chips across Alphabet and extended by external chipmakers. Even so, a non-peer-reviewed invited paper at ISPD 2023 questioned its performance claims, despite failing to run our method as described in Nature. For example, it did not pre-train the RL method (removing its ability to learn from prior experience), used substantially fewer compute resources (20x fewer RL experience collectors and half as many GPUs), did not train to convergence (standard practice in machine learning), and evaluated on test cases that are not representative of modern chips. Recently, Igor Markov published a meta-analysis of three papers: our peer-reviewed Nature paper, the non-peer-reviewed ISPD paper, and Markov's own unpublished paper (though he does not disclose that he co-authored it). Although AlphaChip has already achieved widespread adoption and impact, we publish this response to ensure that no one is wrongly discouraged from innovating in this impactful area.

Training agents that are robust to environmental changes remains a significant challenge in deep reinforcement learning (RL). Unsupervised environment design (UED) has recently emerged to address this issue by generating a set of training environments tailored to the agent's capabilities. While prior works demonstrate that UED has the potential to learn a robust policy, their performance is constrained by the capabilities of the environment generation. To this end, we propose a novel UED algorithm, adversarial environment design via regret-guided diffusion models (ADD). The proposed method guides the diffusion-based environment generator with the regret of the agent to produce environments that the agent finds challenging but conducive to further improvement. By exploiting the representation power of diffusion models, ADD can directly generate adversarial environments while maintaining the diversity of training environments, enabling the agent to effectively learn a robust policy. Our experimental results demonstrate that the proposed method successfully generates an instructive curriculum of environments, outperforming UED baselines in zero-shot generalization across novel, out-of-distribution environments. Project page: https://rllab-snu.github.io/projects/ADD

Policy search methods are crucial in reinforcement learning, offering a framework to address continuous state-action and partially observable problems. However, the complexity of exploring vast policy spaces can lead to significant inefficiencies. Reducing the policy space through policy compression emerges as a powerful, reward-free approach to accelerate the learning process. This technique condenses the policy space into a smaller, representative set while maintaining most of the original effectiveness. Our research focuses on determining the necessary sample size to learn this compressed set accurately. We employ R\'enyi divergence to measure the similarity between true and estimated policy distributions, establishing error bounds for good approximations. To simplify the analysis, we employ the $l_1$ norm, determining sample size requirements for both model-based and model-free settings. Finally, we correlate the error bounds from the $l_1$ norm with those from R\'enyi divergence, distinguishing between policies near the vertices and those in the middle of the policy space, to determine the lower and upper bounds for the required sample sizes.

Batch reinforcement learning enables policy learning without direct interaction with the environment during training, relying exclusively on previously collected sets of interactions. This approach is, therefore, well-suited for high-risk and cost-intensive applications, such as industrial control. Learned policies are commonly restricted to act in a similar fashion as observed in the batch. In a real-world scenario, learned policies are deployed in the industrial system, inevitably leading to the collection of new data that can subsequently be added to the existing recording. The process of learning and deployment can thus take place multiple times throughout the lifespan of a system. In this work, we propose to exploit this iterative nature of applying offline reinforcement learning to guide learned policies towards efficient and informative data collection during deployment, leading to continuous improvement of learned policies while remaining within the support of collected data. We present an algorithmic methodology for iterative batch reinforcement learning based on ensemble-based model-based policy search, augmented with safety and, importantly, a diversity criterion.