In cooperative multi-agent reinforcement learning (MARL), well-designed communication protocols can effectively facilitate consensus among agents, thereby enhancing task performance. Moreover, in large-scale multi-agent systems commonly found in real-world applications, effective communication plays an even more critical role due to the escalated challenge of partial observability compared to smaller-scale setups. In this work, we endeavor to develop a scalable communication protocol for MARL. Unlike previous methods that focus on selecting optimal pairwise communication links-a task that becomes increasingly complex as the number of agents grows-we adopt a global perspective on communication topology design. Specifically, we propose utilizing the exponential topology to enable rapid information dissemination among agents by leveraging its small-diameter and small-size properties. This approach leads to a scalable communication protocol, named ExpoComm. To fully unlock the potential of exponential graphs as communication topologies, we employ memory-based message processors and auxiliary tasks to ground messages, ensuring that they reflect global information and benefit decision-making. Extensive experiments on large-scale cooperative benchmarks, including MAgent and Infrastructure Management Planning, demonstrate the superior performance and robust zero-shot transferability of ExpoComm compared to existing communication strategies. The code is publicly available at https://github.com/LXXXXR/ExpoComm.

The Hegselmann-Krause (HK) model of opinion dynamics describes how opinions held by individuals in a community change over time in response to the opinions of others and their access to the true value, T, to which these opinions relate. Here, I extend the simple HK model to incorporate an Artificially Intelligent (AI) Oracle that averages the opinions of members of the community. Agent-based simulations show that (1) if individuals only have access to the Oracle (and not T), and incorporate the Oracle's opinion as they update their opinions, then all opinions will converge on a common value; (2) in contrast, if all individuals also have access to T, then all opinions will ultimately converge to T, but the presence of an Oracle may delay the time to convergence; (3) if only some individuals have access to T, opinions may not converge to T, but under certain conditions, universal access to the Oracle will guarantee convergence to T; and (4) whether or not the Oracle only accesses the opinions of individuals who have access to T, or whether it accesses the opinions of everyone in the community, makes no marked difference to the extent to which the average opinion differs from T.

Driven by inherent uncertainty and the sim-to-real gap, robust reinforcement learning (RL) seeks to improve resilience against the complexity and variability in agent-environment sequential interactions. Despite the existence of a large number of RL benchmarks, there is a lack of standardized benchmarks for robust RL. Current robust RL policies often focus on a specific type of uncertainty and are evaluated in distinct, one-off environments. In this work, we introduce Robust-Gymnasium, a unified modular benchmark designed for robust RL that supports a wide variety of disruptions across all key RL components-agents' observed state and reward, agents' actions, and the environment. Offering over sixty diverse task environments spanning control and robotics, safe RL, and multi-agent RL, it provides an open-source and user-friendly tool for the community to assess current methods and foster the development of robust RL algorithms. In addition, we benchmark existing standard and robust RL algorithms within this framework, uncovering significant deficiencies in each and offering new insights.

Figure 1: In a challenging real-world experiment with limited communication (agents can only exchange information every 30 seconds), our method RAMEN enables each turtlebot to successfully map the full scene while only physically visiting half of the scene (explored areas and trajectories are colored accordingly). Our method achieves accuracy comparable to the ground truth while the baseline method (DiNNO) fails to converge. Abstract --Multi-agent neural implicit mapping allows robots to collaboratively capture and reconstruct complex environments with high fidelity. However, existing approaches often rely on synchronous communication, which is impractical in real-world scenarios with limited bandwidth and potential communication interruptions. This paper introduces RAMEN: Real-time Asynchronous Multi-agEnt Neural implicit mapping, a novel approach designed to address this challenge. RAMEN employs an uncertainty-weighted multi-agent consensus optimization algorithm that accounts for communication disruptions. When communication is lost between a pair of agents, each agent retains only an outdated copy of its neighbor's map, with the uncertainty of this copy increasing over time since the last communication. Using gradient update information, we quantify the uncertainty associated with each parameter of the neural network map. Neural network maps from different agents are brought to consensus on the basis of their levels of uncertainty, with consensus biased towards network parameters with lower uncertainty. T o achieve this, we derive a weighted variant of the decentralized consensus alternating direction method of multipliers (C-ADMM) algorithm, facilitating robust collaboration among agents with varying communication and update frequencies.

This is a preprint version. Figure 1: We present an effective and efficient coverage path planning approach that exploits a robot manipulator's redundancy and task tolerances to minimize joint space costs. This task has (B) rotational redundancy around the tool's principal axis and (C) translational tolerance tangential to the wok surface, as the finishing disk can have multiple contact points with the wok. Due to the redundancy, infinite possible motions can cover the surface, and our approach finds one that minimizes joint space costs. Abstract -- Many robotic applications, such as sanding, polishing, wiping and sensor scanning, require a manipulator to dexterously cover a surface using its end-effector . In this paper, we provide an efficient and effective coverage path planning approach that leverages a manipulator's redundancy and task tolerances to minimize costs in joint space. We formulate the problem as a Generalized Traveling Salesman Problem and hierarchically streamline the graph size. Our strategy is to identify guide paths that roughly cover the surface and accelerate the computation by solving a sequence of smaller problems.

Multi-agent debate - multiple instances of large language models discussing problems in turn-based interaction - has shown promise for solving knowledge and reasoning tasks. However, these methods show limitations, particularly when scaling them to longer reasoning chains. In this study, we unveil a new issue of multi-agent debate: discussions drift away from the initial problem over multiple turns. We define this phenomenon as problem drift and quantify its presence across ten tasks (i.e., three generative, three knowledge, three reasoning, and one instruction-following task). To identify the reasons for this issue, we perform a human study with eight experts on discussions suffering from problem drift, who find the most common issues are a lack of progress (35% of cases), low-quality feedback (26% of cases), and a lack of clarity (25% of cases). To systematically address the issue of problem drift, we propose DRIFTJudge, a method based on LLM-as-a-judge, to detect problem drift at test-time. We further propose DRIFTPolicy, a method to mitigate 31% of problem drift cases. Our study can be seen as a first step to understanding a key limitation of multi-agent debate, highlighting pathways for improving their effectiveness in the future.

Multiagent Reinforcement Learning (MARL) poses significant challenges due to the exponential growth of state and action spaces and the non-stationary nature of multiagent environments. This results in notable sample inefficiency and hinders generalization across diverse tasks. The complexity is further pronounced in relational settings, where domain knowledge is crucial but often underutilized by existing MARL algorithms. To overcome these hurdles, we propose integrating relational planners as centralized controllers with efficient state abstractions and reinforcement learning. This approach proves to be sample-efficient and facilitates effective task transfer and generalization.

Social media platforms frequently impose restrictive policies to moderate user content, prompting the emergence of creative evasion language strategies. This paper presents a multi-agent framework based on Large Language Models (LLMs) to simulate the iterative evolution of language strategies under regulatory constraints. In this framework, participant agents, as social media users, continuously evolve their language expression, while supervisory agents emulate platform-level regulation by assessing policy violations. To achieve a more faithful simulation, we employ a dual design of language strategies (constraint and expression) to differentiate conflicting goals and utilize an LLM-driven GA (Genetic Algorithm) for the selection, mutation, and crossover of language strategies. The framework is evaluated using two distinct scenarios: an abstract password game and a realistic simulated illegal pet trade scenario. Experimental results demonstrate that as the number of dialogue rounds increases, both the number of uninterrupted dialogue turns and the accuracy of information transmission improve significantly. Furthermore, a user study with 40 participants validates the real-world relevance of the generated dialogues and strategies. Moreover, ablation studies validate the importance of the GA, emphasizing its contribution to long-term adaptability and improved overall results.

Differential Diagnosis (DDx) is a fundamental yet complex aspect of clinical decision-making, in which physicians iteratively refine a ranked list of possible diseases based on symptoms, antecedents, and medical knowledge. While recent advances in large language models have shown promise in supporting DDx, existing approaches face key limitations, including single-dataset evaluations, isolated optimization of components, unrealistic assumptions about complete patient profiles, and single-attempt diagnosis. We introduce a Modular Explainable DDx Agent (MEDDxAgent) framework designed for interactive DDx, where diagnostic reasoning evolves through iterative learning, rather than assuming a complete patient profile is accessible. MEDDxAgent integrates three modular components: (1) an orchestrator (DDxDriver), (2) a history taking simulator, and (3) two specialized agents for knowledge retrieval and diagnosis strategy. To ensure robust evaluation, we introduce a comprehensive DDx benchmark covering respiratory, skin, and rare diseases. We analyze single-turn diagnostic approaches and demonstrate the importance of iterative refinement when patient profiles are not available at the outset. Our broad evaluation demonstrates that MEDDxAgent achieves over 10% accuracy improvements in interactive DDx across both large and small LLMs, while offering critical explainability into its diagnostic reasoning process.

As AI agents are increasingly adopted to collaborate on complex objectives, ensuring the security of autonomous multi-agent systems becomes crucial. We develop simulations of agents collaborating on shared objectives to study these security risks and security trade-offs. We focus on scenarios where an attacker compromises one agent, using it to steer the entire system toward misaligned outcomes by corrupting other agents. In this context, we observe infectious malicious prompts - the multi-hop spreading of malicious instructions. To mitigate this risk, we evaluated several strategies: two "vaccination" approaches that insert false memories of safely handling malicious input into the agents' memory stream, and two versions of a generic safety instruction strategy. While these defenses reduce the spread and fulfillment of malicious instructions in our experiments, they tend to decrease collaboration capability in the agent network. Our findings illustrate potential trade-off between security and collaborative efficiency in multi-agent systems, providing insights for designing more secure yet effective AI collaborations.