More Capable, Less Cooperative? When LLMs Fail At Zero-Cost Collaboration

• URL: http://arxiv.org/abs/2604.07821v1
• Date: Thu, 09 Apr 2026 05:24:27 GMT
• Title: More Capable, Less Cooperative? When LLMs Fail At Zero-Cost Collaboration
• Abstract summary: We study cooperative behavior in a frictionless environment, removing all strategic complexity from cooperation.<n>We separate cooperation failures from competence failures, tracing their origins through agent reasoning analysis.<n>Our findings suggest that scaling intelligence alone will not solve coordination problems in multi-agent systems.
• Score: 1.4254497466846006
• License: http://creativecommons.org/licenses/by/4.0/
• Abstract: Large language model (LLM) agents increasingly coordinate in multi-agent systems, yet we lack an understanding of where and why cooperation failures may arise. In many real-world coordination problems, from knowledge sharing in organizations to code documentation, helping others carries negligible personal cost while generating substantial collective benefits. However, whether LLM agents cooperate when helping neither benefits nor harms the helper, while being given explicit instructions to do so, remains unknown. We build a multi-agent setup designed to study cooperative behavior in a frictionless environment, removing all strategic complexity from cooperation. We find that capability does not predict cooperation: OpenAI o3 achieves only 17% of optimal collective performance while OpenAI o3-mini reaches 50%, despite identical instructions to maximize group revenue. Through a causal decomposition that automates one side of agent communication, we separate cooperation failures from competence failures, tracing their origins through agent reasoning analysis. Testing targeted interventions, we find that explicit protocols double performance for low-competence models, and tiny sharing incentives improve models with weak cooperation. Our findings suggest that scaling intelligence alone will not solve coordination problems in multi-agent systems and will require deliberate cooperative design, even when helping others costs nothing.

Related papers

• Benchmarking Open-Ended Multi-Agent Coordination in Language Agents [48.1363632826625]
  Alem embeds procedurally generated coordination tasks, soft specialisation, communication, and controllable coordination difficulty into a long-horizon survival world.<n>We introduce $alem$, a JAX-based benchmark for open-ended multi-agent coordination built on Craftax-like dynamics.
  arXiv  Detail & Related papers  (2026-06-06T21:13:43Z)
• Cooperative Profiles Predict Multi-Agent LLM Team Performance in AI for Science Workflows [26.30514689505384]
  We show that game-derived cooperative profiles robustly predict downstream performance in AI-for-Science tasks.<n>Our framework offers a fast and inexpensive diagnostic for screening cooperative fitness before costly multi-agent deployment.
  arXiv  Detail & Related papers  (2026-04-22T15:07:54Z)
• Colosseum: Auditing Collusion in Cooperative Multi-Agent Systems [55.51100373104311]
  We present Colosseum, a framework for auditing LLM agents' collusive behavior in multi-agent settings.<n>Colosseum tests each LLM for collusion under different objectives, persuasion tactics, and network topologies.<n>We discover collusion on paper'' when agents plan to collude in text but would often pick non-collusive actions.
  arXiv  Detail & Related papers  (2026-02-16T21:27:38Z)
• CooperBench: Why Coding Agents Cannot be Your Teammates Yet [44.06715229961526]
  CooperBench is a benchmark of over 600 collaborative coding tasks across 12 libraries in 4 programming languages.<n>Agents achieve on average 30% lower success rates when working together compared to performing both tasks individually.<n>Our analysis reveals three key issues: (1) communication channels become jammed with vague, ill-timed, and inaccurate messages; (2) even with effective communication, agents deviate from their commitments; and (3) agents often hold incorrect expectations about others' plans and communication.
  arXiv  Detail & Related papers  (2026-01-19T18:48:37Z)
• The Collaboration Gap [28.553543260404425]
  We propose a collaborative maze-solving benchmark that (i) isolates collaborative capabilities, (ii) modulates problem complexity, (iii) enables scalable automated grading, and (iv) imposes no output constraintsformat.<n>Using this framework, we evaluate 32 leading open- and closed-source models in solo, homogeneous, and heterogeneous pairings.<n>Our results reveal a "collaboration gap": models that perform well solo often degrade substantially when required to collaborate.
  arXiv  Detail & Related papers  (2025-11-04T16:10:57Z)
• Corrupted by Reasoning: Reasoning Language Models Become Free-Riders in Public Goods Games [87.5673042805229]
  How large language models balance self-interest and collective well-being is a critical challenge for ensuring alignment, robustness, and safe deployment.<n>We adapt a public goods game with institutional choice from behavioral economics, allowing us to observe how different LLMs navigate social dilemmas.<n>Surprisingly, we find that reasoning LLMs, such as the o1 series, struggle significantly with cooperation.
  arXiv  Detail & Related papers  (2025-06-29T15:02:47Z)
• Implicitly Aligning Humans and Autonomous Agents through Shared Task Abstractions [42.813774494968214]
  We introduce HA$2$: Hierarchical Ad Hoc Agents, a framework leveraging hierarchical reinforcement learning to mimic the structured approach humans use in collaboration.<n>We evaluate HA$2$ in the Overcooked environment, demonstrating statistically significant improvement over existing baselines when paired with both unseen agents and humans.
  arXiv  Detail & Related papers  (2025-05-07T17:19:17Z)
• Collaborative Gym: A Framework for Enabling and Evaluating Human-Agent Collaboration [50.657070334404835]
  Collaborative Gym is a framework enabling asynchronous, tripartite interaction among agents, humans, and task environments.<n>We instantiate Co-Gym with three representative tasks in both simulated and real-world conditions.<n>Our findings reveal that collaborative agents consistently outperform their fully autonomous counterparts in task performance.
  arXiv  Detail & Related papers  (2024-12-20T09:21:15Z)
• Deconstructing Cooperation and Ostracism via Multi-Agent Reinforcement Learning [3.3751859064985483]
  We show that network rewiring facilitates mutual cooperation even when one agent always offers cooperation. We also find that ostracism alone is not sufficient to make cooperation emerge. Our findings provide insights into the conditions and mechanisms necessary for the emergence of cooperation.
  arXiv  Detail & Related papers  (2023-10-06T23:18:55Z)
• Building Cooperative Embodied Agents Modularly with Large Language Models [104.57849816689559]
  We address challenging multi-agent cooperation problems with decentralized control, raw sensory observations, costly communication, and multi-objective tasks instantiated in various embodied environments. We harness the commonsense knowledge, reasoning ability, language comprehension, and text generation prowess of LLMs and seamlessly incorporate them into a cognitive-inspired modular framework. Our experiments on C-WAH and TDW-MAT demonstrate that CoELA driven by GPT-4 can surpass strong planning-based methods and exhibit emergent effective communication.
  arXiv  Detail & Related papers  (2023-07-05T17:59:27Z)
• Learning Reward Machines in Cooperative Multi-Agent Tasks [75.79805204646428]
  This paper presents a novel approach to Multi-Agent Reinforcement Learning (MARL) It combines cooperative task decomposition with the learning of reward machines (RMs) encoding the structure of the sub-tasks. The proposed method helps deal with the non-Markovian nature of the rewards in partially observable environments.
  arXiv  Detail & Related papers  (2023-03-24T15:12:28Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.