Interpreting Emergent Extreme Events in Multi-Agent Systems

• URL: http://arxiv.org/abs/2601.20538v1
• Date: Wed, 28 Jan 2026 12:32:16 GMT
• Title: Interpreting Emergent Extreme Events in Multi-Agent Systems
• Authors: Ling Tang, Jilin Mei, Dongrui Liu, Chen Qian, Dawei Cheng, Jing Shao, Xia Hu,
• Abstract summary: This paper proposes the first framework for explaining emergent extreme events in multi-agent systems.<n> Specifically, we adapt the Shapley value to faithfully attribute the occurrence of extreme events to each action taken by agents.<n>We then aggregate the attribution scores along the dimensions of time, agent, and behavior to quantify the risk contribution of each dimension.
• Score: 68.6629534508585
• License: http://creativecommons.org/licenses/by/4.0/
• Abstract: Large language model-powered multi-agent systems have emerged as powerful tools for simulating complex human-like systems. The interactions within these systems often lead to extreme events whose origins remain obscured by the black box of emergence. Interpreting these events is critical for system safety. This paper proposes the first framework for explaining emergent extreme events in multi-agent systems, aiming to answer three fundamental questions: When does the event originate? Who drives it? And what behaviors contribute to it? Specifically, we adapt the Shapley value to faithfully attribute the occurrence of extreme events to each action taken by agents at different time steps, i.e., assigning an attribution score to the action to measure its influence on the event. We then aggregate the attribution scores along the dimensions of time, agent, and behavior to quantify the risk contribution of each dimension. Finally, we design a set of metrics based on these contribution scores to characterize the features of extreme events. Experiments across diverse multi-agent system scenarios (economic, financial, and social) demonstrate the effectiveness of our framework and provide general insights into the emergence of extreme phenomena.

Related papers

• The Why Behind the Action: Unveiling Internal Drivers via Agentic Attribution [63.61358761489141]
  Large Language Model (LLM)-based agents are widely used in real-world applications such as customer service, web navigation, and software engineering.<n>We propose a novel framework for textbfgeneral agentic attribution, designed to identify the internal factors driving agent actions regardless of the task outcome.<n>We validate our framework across a diverse suite of agentic scenarios, including standard tool use and subtle reliability risks like memory-induced bias.
  arXiv  Detail & Related papers  (2026-01-21T15:22:21Z)
• MACIE: Multi-Agent Causal Intelligence Explainer for Collective Behavior Understanding [0.2538209532048867]
  We present MACIE, a framework combining structural causal models, interventional counterfactuals, and Shapley values to provide comprehensive explanations.<n>We evaluate MACIE across four MARL scenarios: cooperative, competitive, and mixed motive.<n>Results show accurate outcome attribution, mean phi_i equals 5.07, standard deviation less than 0.05, detection of positive emergence in cooperative tasks, synergy index up to 0.461, and efficient computation, 0.79 seconds per dataset on CPU.
  arXiv  Detail & Related papers  (2025-11-11T19:04:22Z)
• DeceptionBench: A Comprehensive Benchmark for AI Deception Behaviors in Real-world Scenarios [57.327907850766785]
  characterization of deception across realistic real-world scenarios remains underexplored.<n>We establish DeceptionBench, the first benchmark that systematically evaluates how deceptive tendencies manifest across different domains.<n>On the intrinsic dimension, we explore whether models exhibit self-interested egoistic tendencies or sycophantic behaviors that prioritize user appeasement.<n>We incorporate sustained multi-turn interaction loops to construct a more realistic simulation of real-world feedback dynamics.
  arXiv  Detail & Related papers  (2025-10-17T10:14:26Z)
• Timescale-agnostic characterisation for collective attention events [0.0]
  We find four characteristic behaviours in the usage of hashtags on Twitter that are indicative of different patterns of attention to topics. We develop an agent-based model for generating collective attention events to test the factors affecting emergence of these phenomena.
  arXiv  Detail & Related papers  (2024-11-18T12:01:59Z)
• Episodic Future Thinking Mechanism for Multi-agent Reinforcement Learning [2.992602379681373]
  We introduce an episodic future thinking (EFT) mechanism for a reinforcement learning (RL) agent. We first develop a multi-character policy that captures diverse characters with an ensemble of heterogeneous policies. Once the character is inferred, the agent predicts the upcoming actions of target agents and simulates the potential future scenario.
  arXiv  Detail & Related papers  (2024-10-22T19:12:42Z)
• BattleAgent: Multi-modal Dynamic Emulation on Historical Battles to Complement Historical Analysis [62.60458710368311]
  This paper presents BattleAgent, an emulation system that combines the Large Vision-Language Model and Multi-agent System. It aims to simulate complex dynamic interactions among multiple agents, as well as between agents and their environments. It emulates both the decision-making processes of leaders and the viewpoints of ordinary participants, such as soldiers.
  arXiv  Detail & Related papers  (2024-04-23T21:37:22Z)
• Clustering-based Identification of Precursors of Extreme Events in Chaotic Systems [0.0]
  Abrupt and rapid high-amplitude changes in a dynamical system's states known as extreme event appear in many processes occurring in nature. The applicability of a data-driven modularity-based clustering technique to identify precursors of rare and extreme events in chaotic systems is explored.
  arXiv  Detail & Related papers  (2023-06-20T12:38:38Z)
• Modeling Complex Event Scenarios via Simple Entity-focused Questions [58.16787028844743]
  We propose a question-guided generation framework that models events in complex scenarios as answers to questions about participants. At any step in the generation process, the framework uses the previously generated events as context, but generates the next event as an answer to one of three questions. Our empirical evaluation shows that this question-guided generation provides better coverage of participants, diverse events within a domain, and comparable perplexities for modeling event sequences.
  arXiv  Detail & Related papers  (2023-02-14T15:48:56Z)
• Beyond Grounding: Extracting Fine-Grained Event Hierarchies Across Modalities [43.048896440009784]
  We propose the task of extracting event hierarchies from multimodal (video and text) data. This reveals the structure of events and is critical to understanding them. We show the limitations of state-of-the-art unimodal and multimodal baselines on this task.
  arXiv  Detail & Related papers  (2022-06-14T23:24:15Z)
• CausalCity: Complex Simulations with Agency for Causal Discovery and Reasoning [68.74447489372037]
  We present a high-fidelity simulation environment that is designed for developing algorithms for causal discovery and counterfactual reasoning. A core component of our work is to introduce textitagency, such that it is simple to define and create complex scenarios. We perform experiments with three state-of-the-art methods to create baselines and highlight the affordances of this environment.
  arXiv  Detail & Related papers  (2021-06-25T00:21:41Z)
• Modelling Cooperation in Network Games with Spatio-Temporal Complexity [11.665246332943058]
  We study the emergence of self-organized cooperation in complex gridworld domains. Using multi-agent deep reinforcement learning, we simulate an agent society for a variety of plausible mechanisms. Our methods have implications for mechanism design in both human and artificial agent systems.
  arXiv  Detail & Related papers  (2021-02-13T12:04:52Z)

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.