Autonomous Agents and Multi-Agent Systems最新文献

IF 2.4 3区计算机科学

Autonomous Agents and Multi-Agent Systems Pub Date : 2026-10-13 DOI: 10.1007/s10458-026-09775-3

Laurent Gourvès, Michael Lampis, Nikolaos Melissinos, Aris Pagourtzis

{"title":"Satisfactory budget division","authors":"Laurent Gourvès, Michael Lampis, Nikolaos Melissinos, Aris Pagourtzis","doi":"10.1007/s10458-026-09775-3","DOIUrl":"10.1007/s10458-026-09775-3","url":null,"abstract":"<div>A divisible budget must be allocated to several projects, and agents are asked for their opinion on how much they would give to each project. We consider that an agent is satisfied by a division of the budget if, for at least a certain predefined number (tau ) of projects, the part of the budget actually allocated to each project is at least as large as the amount the agent requested. The objective is to find a budget division that “best satisfies” the agents. In this context, different problems can be stated and we address the following ones. We study (i) the largest proportion of agents that can be satisfied for any instance, (ii) classes of instances admitting a budget division that satisfies all agents, (iii) the complexity of deciding if, for a given instance, every agent can be satisfied, and finally (iv) the question of finding, for a given instance, the smallest total budget to satisfy all agents. We provide answers to these complementary questions for several natural values of the parameter (tau ), capturing scenarios where we seek to satisfy for each agent all; almost all; half; or at least one of her requests.</div>","PeriodicalId":55586,"journal":{"name":"Autonomous Agents and Multi-Agent Systems","volume":"40 2","pages":""},"PeriodicalIF":2.4,"publicationDate":"2026-10-13","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"149189104","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Rectifying market externalities via AI policymaking 通过人工智能决策纠正市场外部性

IF 2.4 3区计算机科学

Autonomous Agents and Multi-Agent Systems Pub Date : 2026-09-30 DOI: 10.1007/s10458-026-09768-2

Panayiotis Danassis, Aris Filos-Ratsikas, Haipeng Chen, Milind Tambe, Boi Faltings

引用次数: 0

Curiosity-driven partner selection accelerates convention emergence in language games 好奇心驱动的合作伙伴选择加速了语言游戏中的惯例形成

IF 2.4 3区计算机科学

Autonomous Agents and Multi-Agent Systems Pub Date : 2026-09-16 DOI: 10.1007/s10458-026-09771-7

Chin-wing Leung, Paolo Turrini, Ann Nowé

{"title":"Curiosity-driven partner selection accelerates convention emergence in language games","authors":"Chin-wing Leung, Paolo Turrini, Ann Nowé","doi":"10.1007/s10458-026-09771-7","DOIUrl":"10.1007/s10458-026-09771-7","url":null,"abstract":"<div>In language games a speaker and a listener attempt to coordinate on a shared mapping between words and concepts. The usual approach in the literature is to study convention emergence in well-mixed populations, where pairs of agents are randomly matched to play the role of speaker and listener, respectively. This way of pairing agents can be shown to promote the emergence of a unifying common language in the long run. Despite the theoretical guarantee, convention emergence can be very slow and practically infeasible, especially in large populations with many words and concepts. Here, we propose an alternative approach, where we allow agents to selectively partner with other agents based on their past experience. To this aim, we study Boltzmann Q-learning agents that are curiosity-driven, i.e., more likely to choose partners they misunderstood in the past. We show that this selection method significantly accelerates convention emergence when compared against a random-matching baseline and is even more pronounced when restricting agents’ communication channels or considering an imbalanced number of concepts and words. By inspecting the evolution of the agents’ interaction frequency we see that partner selection induces low treewidth and high degree variance at the early stages of learning, to then converge to a regular graph, which allows for settling misunderstandings in the population at a faster rate than the traditional approaches.</div>","PeriodicalId":55586,"journal":{"name":"Autonomous Agents and Multi-Agent Systems","volume":"40 2","pages":""},"PeriodicalIF":2.4,"publicationDate":"2026-09-16","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://link.springer.com/content/pdf/10.1007/s10458-026-09771-7.pdf","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148985357","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Contiguous allocation of indivisible items on a path 路径上不可分割的项目的连续分配

IF 2.4 3区计算机科学

Autonomous Agents and Multi-Agent Systems Pub Date : 2026-09-15 DOI: 10.1007/s10458-026-09773-5

Yasushi Kawase, Bodhayan Roy, Mohammad Azharuddin Sanpui

{"title":"Contiguous allocation of indivisible items on a path","authors":"Yasushi Kawase, Bodhayan Roy, Mohammad Azharuddin Sanpui","doi":"10.1007/s10458-026-09773-5","DOIUrl":"10.1007/s10458-026-09773-5","url":null,"abstract":"<div>We study the problem of allocating indivisible items on a path among agents. The objective is to find a fair and efficient allocation in which each agent’s bundle forms a contiguous block on the line. We examine two natural settings: the fixed-order setting, where the order is predetermined, and the flexible-order setting, where the sequence of agents is unrestricted. In the fixed-order setting, we show that both maximum utilitarian social welfare and egalitarian social welfare can be computed in polynomial time. Moreover, allocations satisfying proportionality, equitability, and the maximin share guarantee can also be computed in polynomial time in this setting. In contrast, checking the existence of an EF1 allocation is NP-complete for binary additive valuations. In the flexible-order setting, we demonstrate that, even when the valuations are binary additive, deciding whether every item can be allocated to an agent who values it is NP-complete. Consequently, we provide two fixed-parameter tractable (FPT) algorithms for maximizing utilitarian social welfare, parameterized by the number of agents and the number of items. Additionally, we present a (varvec{1/2})-approximation algorithm for a special case. Furthermore, we establish that deciding whether the maximum egalitarian social welfare is at least (varvec{2}) or at most (varvec{1}) is NP-complete for binary additive valuations. We also give two FPT algorithms for maximizing egalitarian social welfare, parameterized by the number of agents and the number of items.</div>","PeriodicalId":55586,"journal":{"name":"Autonomous Agents and Multi-Agent Systems","volume":"40 2","pages":""},"PeriodicalIF":2.4,"publicationDate":"2026-09-15","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148986446","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Distributed multi-agent coordination for energy-constrained patrolling 能量约束下巡逻的分布式多智能体协调

IF 2.4 3区计算机科学

Autonomous Agents and Multi-Agent Systems Pub Date : 2026-09-09 DOI: 10.1007/s10458-026-09764-6

Brendan Sims, Zakria Qadir, Haris Aziz

{"title":"Distributed multi-agent coordination for energy-constrained patrolling","authors":"Brendan Sims, Zakria Qadir, Haris Aziz","doi":"10.1007/s10458-026-09764-6","DOIUrl":"10.1007/s10458-026-09764-6","url":null,"abstract":"<div>In this paper, the multi-agent patrolling problem subject to energy constraints is studied. Several existing online, distributed, heuristic-based patrolling algorithms are adapted to accommodate energy-constrained multi-agent patrolling. These adapted algorithms are evaluated on graphs using simulations in order to identify algorithm characteristics that are beneficial for energy-constrained multi-agent patrolling and to analyse the performance of the algorithms when subject to energy constraints. The experimental results also reveal several important considerations when planning multi-agent patrolling missions subject to energy constraints. Patrolling performance is measured using metrics based on idleness, which represents the time since a location within a patrolling area was previously visited. It was evident from the experimental results that coordination between agents, such as information sharing and/or conflict resolution, improves performance in general, especially as the number of agents increases. However, in the problem setting considered, sharing visit information alone appears to be detrimental to patrolling performance. Furthermore, careful consideration may be required regarding the patrolling area in order to ensure that agents subject to energy constraints can reach all locations when using certain distributed, heuristic-based patrolling algorithms. In addition, increasing the amount of energy available can negatively impact the performance of distributed, heuristic-based patrolling algorithms on some graphs. Overall, no single algorithm consistently outperformed all others across all graph topologies and different numbers of agents with respect to the evaluated patrolling metrics. This highlights the importance of considering each factor when planning distributed, heuristic-based multi-agent patrolling strategies under energy constraints.</div>","PeriodicalId":55586,"journal":{"name":"Autonomous Agents and Multi-Agent Systems","volume":"40 2","pages":""},"PeriodicalIF":2.4,"publicationDate":"2026-09-09","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://link.springer.com/content/pdf/10.1007/s10458-026-09764-6.pdf","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148926824","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

LLM-guided contrastive role representation for diverse multi-agent coordination 基于llm的多智能体协调对比角色表示

IF 2.4 3区计算机科学

Autonomous Agents and Multi-Agent Systems Pub Date : 2026-09-05 DOI: 10.1007/s10458-026-09769-1

Tingting Wei, Yan Zheng, Zhangling Wang, Chushu Yi, Chengyi Zeng, Lina Lu, Shaofei Chen, Jing Chen

{"title":"LLM-guided contrastive role representation for diverse multi-agent coordination","authors":"Tingting Wei, Yan Zheng, Zhangling Wang, Chushu Yi, Chengyi Zeng, Lina Lu, Shaofei Chen, Jing Chen","doi":"10.1007/s10458-026-09769-1","DOIUrl":"10.1007/s10458-026-09769-1","url":null,"abstract":"<div>Multi-agent reinforcement learning (MARL) has achieved notable progress in cooperative tasks, yet the common use of parameter sharing often leads to behavioral homogeneity, which restricts the diversity and adaptability of agent coordination. To address this issue, we propose Contrastive Learning for Role Representation (CLRE), a framework designed to promote strategic diversity by encouraging agents to learn distinct and semantically meaningful roles. The key idea is to employ a Large Language Model (LLM) as an offline domain expert that automatically generates semantic role labels from agent states. These labels provide reliable supervision for a contrastive learning objective, guiding the model to acquire role embeddings that are both discriminative and interpretable. To ensure these role representations effectively support coordination, CLRE incorporates a graph attention network that captures inter-agent relationships and a dual-attention value function that integrates role information with the global state for context-aware decision-making. Experiments on the StarCraft Multi-Agent Challenge (SMAC) benchmark demonstrate that CLRE achieves superior win rates and sample efficiency compared with state-of-the-art MARL methods. Ablation studies further confirm that each component contributes to the overall performance, showing that the proposed approach enables diverse and efficient cooperation among agents in complex environments.</div>","PeriodicalId":55586,"journal":{"name":"Autonomous Agents and Multi-Agent Systems","volume":"40 2","pages":""},"PeriodicalIF":2.4,"publicationDate":"2026-09-05","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148926751","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Stable marriage in Euclidean space 欧几里得空间里稳定的婚姻

IF 2.4 3区计算机科学

Autonomous Agents and Multi-Agent Systems Pub Date : 2026-09-05 DOI: 10.1007/s10458-026-09767-3

Yinghui Wen, José Rodríguez, Zhongyi Zhang, David Manlove, Jiong Guo

{"title":"Stable marriage in Euclidean space","authors":"Yinghui Wen, José Rodríguez, Zhongyi Zhang, David Manlove, Jiong Guo","doi":"10.1007/s10458-026-09767-3","DOIUrl":"10.1007/s10458-026-09767-3","url":null,"abstract":"<div>We study stable marriage problems in the d-Euclidean space. Under this setting, each agent is represented as a point in the d-dimensional space, and for each agent a, the preference of a is based on the sorting according to the Euclidean distances between a and agents from the opposite gender. Let (delta (a,b)) be the Euclidean distance between two points a and b. A man u prefers a woman (w_1) to another woman (w_2) if and only if (delta (u,w_1) < delta (u,w_2)). If (delta (u,w_1) = delta (u,w_2)), then u ranks (w_1) and (w_2) indifferently, and we say there is a tie between (w_1) and (w_2) in u’s preference list. A lot of variants of Stable Marriage with Ties (SMT) have been shown to be NP-complete when ties occur in preference lists. In this paper, we study the most famous hard variants of SMT in d-Euclidean space, namely, Regret-SMT, Egalitarian-SMT, Forced-SMT, Max-SMTI, and MinStrong-SMT. We prove that, with (d=1), Regret-SMT and Forced-SMT can be solved in polynomial time, and MinStrong-SMT can be solved in polynomial time when all agents occupy pairwise distinct positions. When the number of dimensions is unbounded, all five problems are NP-hard. This result is strengthened with a separate reduction that shows that these five problems are NP-hard even when (d=2). Then we show that if the preference lists can be incomplete (agents are allowed to not give a full rank of the opposite gender), all five problems are NP-hard, even with (d=1). Finally, we provide a polynomial-time algorithm to recognize whether a given preference profile with complete preferences can be embedded into 1-Euclidean space, and prove that this problem becomes NP-hard when preference lists are incomplete.</div>","PeriodicalId":55586,"journal":{"name":"Autonomous Agents and Multi-Agent Systems","volume":"40 2","pages":""},"PeriodicalIF":2.4,"publicationDate":"2026-09-05","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148926755","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Coordinated motion planning: Multi-agent path finding in a densely packed, bounded domain 协调运动规划：在密集的有界域中寻找多智能体路径

IF 2.4 3区计算机科学

Autonomous Agents and Multi-Agent Systems Pub Date : 2026-09-02 DOI: 10.1007/s10458-026-09766-4

Sándor P. Fekete, Ramin Kosfeld, Peter Kramer, Jonas Neutzner, Christian Rieck, Christian Scheffer

{"title":"Coordinated motion planning: Multi-agent path finding in a densely packed, bounded domain","authors":"Sándor P. Fekete, Ramin Kosfeld, Peter Kramer, Jonas Neutzner, Christian Rieck, Christian Scheffer","doi":"10.1007/s10458-026-09766-4","DOIUrl":"10.1007/s10458-026-09766-4","url":null,"abstract":"<div>We study Multi-Agent Path Finding for arrangements of labeled agents in the interior of a simply connected domain: Given a unique start and target position for each agent, the goal is to find a sequence of parallel, collision-free agent motions that minimizes the overall time (the makespan) until all agents have reached their respective targets. A natural case is that of a simply connected polygonal domain with axis-parallel boundaries and integer coordinates, i.e., a simple polyomino, which amounts to a simply connected union of lattice unit squares or cells. We focus on the particularly challenging setting of densely packed agents, i.e., one per cell, which strongly restricts the mobility of agents, and requires intricate coordination of motion. We provide a variety of novel results for this problem, including (1) a characterization of polyominoes in which a reconfiguration plan is guaranteed to exist; (2) a characterization of shape parameters that induce worst-case bounds on the makespan; (3) a suite of algorithms to achieve asymptotically worst-case optimal performance with respect to the achievable stretch for cases with severely limited maneuverability. This corresponds to bounding the ratio between obtained makespan and the lower bound provided by the max-min distance between the start and target position of any agent and our shape parameters. Our results extend findings by Demaine et al. [18, 19] who investigated the problem for solid rectangular domains, and in the closely related field of Permutation Routing, as presented by Alpert et al. [10] for convex pieces of grid graphs.</div>","PeriodicalId":55586,"journal":{"name":"Autonomous Agents and Multi-Agent Systems","volume":"40 2","pages":""},"PeriodicalIF":2.4,"publicationDate":"2026-09-02","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://link.springer.com/content/pdf/10.1007/s10458-026-09766-4.pdf","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148926704","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Consensus from Ambiguity: Language Emergence in Guessing Games 歧义的共识：猜谜游戏中的语言涌现

IF 2.4 3区计算机科学

Autonomous Agents and Multi-Agent Systems Pub Date : 2026-08-24 DOI: 10.1007/s10458-026-09762-8

Ajin George Joseph, Vikas Kumar

引用次数: 0

Whom to trust? selective online learning in multi-agent systems with prior-aware gaussian process regression 该信任谁？基于先验感知高斯过程回归的多智能体系统的选择性在线学习

IF 2.4 3区计算机科学

Autonomous Agents and Multi-Agent Systems Pub Date : 2026-08-14 DOI: 10.1007/s10458-026-09765-5

Zewen Yang, Xiaobing Dai, Akshat Dubey, Yansong Wu, Xiao Chen, Ebenezer Oladimeji Awotoro, Sandra Hirche

{"title":"Whom to trust? selective online learning in multi-agent systems with prior-aware gaussian process regression","authors":"Zewen Yang, Xiaobing Dai, Akshat Dubey, Yansong Wu, Xiao Chen, Ebenezer Oladimeji Awotoro, Sandra Hirche","doi":"10.1007/s10458-026-09765-5","DOIUrl":"10.1007/s10458-026-09765-5","url":null,"abstract":"<div>This paper introduces a novel approach to advancing distributed cooperative learning in multi-agent systems (MASs) using Gaussian Process (GP) regression. The primary contribution lies in the development of a prior-aware selective distributed GPs (Pri-GP) algorithm, which allows agents to selectively collaborate by requesting predictions from neighboring agents based on their trustworthiness. This selective learning mechanism ensures that agents can avoid relying on neighbors with potentially flawed prior knowledge, thereby improving prediction accuracy, particularly in scenarios where an agent’s prior information is incorrect or outdated. Moreover, Pri-GP eliminates the need for costly variance calculations typically required for determining aggregation weights. We also extend this framework to the online-Pri-GP algorithm, which integrates real-time learning strategies, demonstrating superior performance in dynamical systems by achieving lower prediction errors. Additionally, we provide a prediction error bound within the Pri-GP framework, offering a guarantee of the aggregated predictions. Simulation results demonstrate that Pri-GP and online-Pri-GP outperform existing distributed GPs approaches in function approximation and system identification.</div>","PeriodicalId":55586,"journal":{"name":"Autonomous Agents and Multi-Agent Systems","volume":"40 2","pages":""},"PeriodicalIF":2.4,"publicationDate":"2026-08-14","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148751812","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0