One part of complying with norms, rules, and preferences is incorporating constraints (such as knowledge of ethics) into one's goal formulation and planning processing. We explore in a simple domain how the encoding of knowledge in different ethical frameworks influences an agent's goal formulation and planning processing and demonstrate ability of an agent to satisfy and satisfice when its collection of relevant constraints includes a mix of "hard" and "soft" constraints of various types. How the agent attempts to comply with ethical constraints depends on the ethical framing and we investigate tradeoffs between deontological framing and utilitarian framing for complying with an ethical norm. Representative scenarios highlight how performing the same task with different framings of the same norm leads to different behaviors. Our explorations suggest an important role for metacognitive judgments in resolving ethical conflicts during goal formulation and planning.

To enable effective human-AI collaboration, merely optimizing AI performance while ignoring humans is not sufficient. Recent research has demonstrated that designing AI agents to account for human behavior leads to improved performance in human-AI collaboration. However, a limitation of most existing approaches is their assumption that human behavior is static, irrespective of AI behavior. In reality, humans may adjust their action plans based on their observations of AI behavior. In this paper, we address this limitation by enabling a collaborative AI agent to consider the beliefs of its human partner, i.e., what the human partner thinks the AI agent is doing, and design its action plan to facilitate easier collaboration with its human partner. Specifically, we developed a model of human beliefs that accounts for how humans reason about the behavior of their AI partners. Based on this belief model, we then developed an AI agent that considers both human behavior and human beliefs in devising its strategy for working with humans. Through extensive real-world human-subject experiments, we demonstrated that our belief model more accurately predicts humans' beliefs about AI behavior. Moreover, we showed that our design of AI agents that accounts for human beliefs enhances performance in human-AI collaboration.

Many multi-agent systems in practice are decentralized and have dynamically varying dependencies. There has been a lack of attempts in the literature to analyze these systems theoretically. In this paper, we propose and theoretically analyze a decentralized model with dynamically varying dependencies called the Locally Interdependent Multi-Agent MDP. This model can represent problems in many disparate domains such as cooperative navigation, obstacle avoidance, and formation control. Despite the intractability that general partially observable multi-agent systems suffer from, we propose three closed-form policies that are theoretically near-optimal in this setting and can be scalable to compute and store. Consequentially, we reveal a fundamental property of Locally Interdependent Multi-Agent MDP's that the partially observable decentralized solution is exponentially close to the fully observable solution with respect to the visibility radius. We then discuss extensions of our closed-form policies to further improve tractability. We conclude by providing simulations to investigate some long horizon behaviors of our closed-form policies.

The rapidly growing capability of Large Language Models (LLMs) is dramatically reshaping many industries [2, 3, 4]. The most recent release of GPT-4o [5] demonstrates a significant leap in multi-modal capabilities and artificial intelligence (AI)-human interaction, whilst maintaining the same level of text generation, reasoning, and code intelligence as GPT-4-Turbo [6]. LLMs can interact with humans and the world as humans do, it is considered a starting point for LLMs to take over tasks from humans or collaborate naturally with humans. Issue resolving is one of the software engineering tasks experimented with LLMs that is particularly relevant in practice. SWE-bench [1] collects 2,294 real-world issues from 12 popular Python libraries.

Autonomous agents that execute human tasks by controlling computers can enhance human productivity and application accessibility. However, progress in this field will be driven by realistic and reproducible benchmarks. We present AndroidWorld, a fully functional Android environment that provides reward signals for 116 programmatic tasks across 20 real-world Android apps. Unlike existing interactive environments, which provide a static test set, AndroidWorld dynamically constructs tasks that are parameterized and expressed in natural language in unlimited ways, thus enabling testing on a much larger and more realistic suite of tasks. Reward signals are derived from the computer's system state, making them durable across task variations and extensible across different apps. To demonstrate AndroidWorld's benefits and mode of operation, we introduce a new computer control agent, M3A. M3A can complete 30.6% of the AndroidWorld's tasks, leaving ample room for future work. Furthermore, we adapt a popular desktop web agent to work on Android, which we find to be less effective on mobile, suggesting future research is needed to achieve universal, cross-domain agents. Finally, we conduct a robustness analysis by testing M3A against a range of task variations on a representative subset of tasks, demonstrating that variations in task parameters can significantly alter a task's complexity and, consequently, an agent's performance, highlighting the importance of testing agents under diverse conditions. AndroidWorld and the experiments in this paper are available at https://github.com/google-research/android_world.

In Multi-agent Reinforcement Learning (MARL), accurately perceiving opponents' strategies is essential for both cooperative and adversarial contexts, particularly within dynamic environments. While Proximal Policy Optimization (PPO) and related algorithms such as Actor-Critic with Experience Replay (ACER), Trust Region Policy Optimization (TRPO), and Deep Deterministic Policy Gradient (DDPG) perform well in single-agent, stationary environments, they suffer from high variance in MARL due to non-stationary and hidden policies of opponents, leading to diminished reward performance. Additionally, existing methods in MARL face significant challenges, including the need for inter-agent communication, reliance on explicit reward information, high computational demands, and sampling inefficiencies. These issues render them less effective in continuous environments where opponents may abruptly change their policies without prior notice. Against this background, we present OPS-DeMo (Online Policy Switch-Detection Model), an online algorithm that employs dynamic error decay to detect changes in opponents' policies. OPS-DeMo continuously updates its beliefs using an Assumed Opponent Policy (AOP) Bank and selects corresponding responses from a pre-trained Response Policy Bank. Each response policy is trained against consistently strategizing opponents, reducing training uncertainty and enabling the effective use of algorithms like PPO in multi-agent environments. Comparative assessments show that our approach outperforms PPO-trained models in dynamic scenarios like the Predator-Prey setting, providing greater robustness to sudden policy shifts and enabling more informed decision-making through precise opponent policy insights.

Human-robot collaboration (HRC) in a shared workspace has become a common pattern in real-world robot applications and has garnered significant research interest. However, most existing studies for human-in-the-loop (HITL) collaboration with robots in a shared workspace evaluate in either simplified game environments or physical platforms, falling short in limited realistic significance or limited scalability. To support future studies, we build an embodied framework named HumanTHOR, which enables humans to act in the simulation environment through VR devices to support HITL collaborations in a shared workspace. To validate our system, we build a benchmark of everyday tasks and conduct a preliminary user study with two baseline algorithms. The results show that the robot can effectively assist humans in collaboration, demonstrating the significance of HRC. The comparison among different levels of baselines affirms that our system can adequately evaluate robot capabilities and serve as a benchmark for different robot algorithms. The experimental results also indicate that there is still much room in the area and our system can provide a preliminary foundation for future HRC research in a shared workspace. More information about the simulation environment, experiment videos, benchmark descriptions, and additional supplementary materials can be found on the website: https://sites.google.com/view/humanthor/.

In this paper, we introduce Multi-Objective Deep Centralized Multi-Agent Actor-Critic (MO- DCMAC), a multi-objective reinforcement learning (MORL) method for infrastructural maintenance optimization, an area traditionally dominated by single-objective reinforcement learning (RL) approaches. Previous single-objective RL methods combine multiple objectives, such as probability of collapse and cost, into a singular reward signal through reward-shaping. In contrast, MO-DCMAC can optimize a policy for multiple objectives directly, even when the utility function is non-linear. We evaluated MO-DCMAC using two utility functions, which use probability of collapse and cost as input. The first utility function is the Threshold utility, in which MO-DCMAC should minimize cost so that the probability of collapse is never above the threshold. The second is based on the Failure Mode, Effects, and Criticality Analysis (FMECA) methodology used by asset managers to asses maintenance plans. We evaluated MO-DCMAC, with both utility functions, in multiple maintenance environments, including ones based on a case study of the historical quay walls of Amsterdam. The performance of MO-DCMAC was compared against multiple rule-based policies based on heuristics currently used for constructing maintenance plans. Our results demonstrate that MO-DCMAC outperforms traditional rule-based policies across various environments and utility functions.

Markov games (MGs) and multi-agent reinforcement learning (MARL) are studied to model decision making in multi-agent systems. Traditionally, the objective in MG and MARL has been risk-neutral, i.e., agents are assumed to optimize a performance metric such as expected return, without taking into account subjective or cognitive preferences of themselves or of other agents. However, ignoring such preferences leads to inaccurate models of decision making in many real-world scenarios in finance, operations research, and behavioral economics. Therefore, when these preferences are present, it is necessary to incorporate a suitable measure of risk into the optimization objective of agents, which opens the door to risk-sensitive MG and MARL. In this paper, we systemically review the literature on risk sensitivity in MG and MARL that has been growing in recent years alongside other areas of reinforcement learning and game theory. We define and mathematically describe different risk measures used in MG and MARL and individually for each measure, discuss articles that incorporate it. Finally, we identify recent trends in theoretical and applied works in the field and discuss possible directions of future research.

For autonomous agents to successfully integrate into human-centered environments, agents should be able to learn from and adapt to humans in their native settings. Preference-based reinforcement learning (PbRL) is a promising approach that learns reward functions from human preferences. This enables RL agents to adapt their behavior based on human desires. However, humans live in a world full of diverse information, most of which is not relevant to completing a particular task. It becomes essential that agents learn to focus on the subset of task-relevant environment features. Unfortunately, prior work has largely ignored this aspect; primarily focusing on improving PbRL algorithms in standard RL environments that are carefully constructed to contain only task-relevant features. This can result in algorithms that may not effectively transfer to a more noisy real-world setting. To that end, this work proposes R2N (Robust-to-Noise), the first PbRL algorithm that leverages principles of dynamic sparse training to learn robust reward models that can focus on task-relevant features. We study the effectiveness of R2N in the Extremely Noisy Environment setting, an RL problem setting where up to 95% of the state features are irrelevant distractions. In experiments with a simulated teacher, we demonstrate that R2N can adapt the sparse connectivity of its neural networks to focus on task-relevant features, enabling R2N to significantly outperform several state-of-the-art PbRL algorithms in multiple locomotion and control environments.