mlsanityData to 14 Oct 2026

Rethinking Failure Attribution in Multi-Agent Systems: A Multi-Perspective Benchmark and Evaluation

arXiv 2603.25001 ↗First posted 26 Mar 2026cs.AIcited by 24

On the map: LLM reasoning → Reasoning for agents, planning and code

also in 1 more
Abstract (arXiv)

Failure attribution is essential for diagnosing and improving multi-agent systems (MAS), yet existing benchmarks and methods largely assume a single deterministic root cause for each failure. In practice, MAS failures often admit multiple plausible attributions due to complex inter-agent dependencies and ambiguous execution trajectories. We revisit MAS failure attribution from a multi-perspective standpoint and propose multi-perspective failure attribution, a practical paradigm that explicitly accounts for attribution ambiguity. To support this setting, we introduce MP-Bench, the first benchmark designed for multi-perspective failure attribution in MAS, along with a new evaluation protocol tailored to this paradigm. Through extensive experiments, we find that prior conclusions suggesting LLMs struggle with failure attribution are largely driven by limitations in existing benchmark designs. Our results highlight the necessity of multi-perspective benchmarks and evaluation protocols for realistic and reliable MAS debugging.

Key facts. Cited by 24 papers in the map (data through 14 Oct 2026). Placed in Reasoning for agents, planning and code, one of 17 directions in LLM reasoning. Builds on 18 of its 34 references that are in the map. First posted on arXiv 26 Mar 2026.

Builds on

18 of 34 references.

Closest ancestors · 2

Same direction, newest first

  1. 2025MAS-GPT: Training LLMs to Build LLM-based Multi-Agent Systems5 Mar
  2. 2024AFlow: Automating Agentic Workflow Generation14 Oct

Other references · 16

  1. 2025Qwen3 Technical Reportcited by 8,820
  2. 2025gpt-oss-120b & gpt-oss-20b Model Cardcited by 1,386
  3. 2025Why Do Multi-Agent LLM Systems Fail?cited by 604
  4. 2024Magentic-One: A Generalist Multi-Agent System for Solving Complex Taskscited by 232
  5. 2024AssistantBench: Can Web Agents Solve Realistic and Time-Consuming Tasks?cited by 158
  6. 2025AgenTracer: Who Is Inducing Failure in the LLM Agentic Systems?cited by 116
  7. 2023When the Majority is Wrong: Modeling Annotator Disagreement for Subjective Taskscited by 90
  8. 2024Adaptive In-conversation Team Building for Language Model Agentscited by 39
  9. 2024The Perspectivist Paradigm Shift: Assumptions and Challenges of Capturing Human Labelscited by 38
  10. 2025Beyond Demographics: Fine-tuning Large Language Models to Predict Individuals' Subjective Text Perceptionscited by 33
  11. 2025GraphTracer: Graph-Guided Failure Tracing in LLM Agents for Robust Multi-Turn Deep Searchcited by 23
  12. 2025Aegis: Automated Error Generation and Attribution for Multi-Agent Systemscited by 22
4 more
  1. 2025Automatic Failure Attribution and Critical Step Prediction Method for Multi-Agent Systems Based on Causal Inferencecited by 18
  2. 2025Abduct, Act, Predict: Scaffolding Causal Inference for Automated Failure Attribution in Multi-Agent Systemscited by 14
  3. 2025Modeling Annotator Disagreement with Demographic-Aware Experts and Synthetic Perspectivescited by 10
  4. 2025Dealing with Annotator Disagreement in Hate Speech Classificationcited by 3

Built on by

Cited by year (2026 to 14 Oct)

Citing papers per year: 2026: 24.

24 papers.

Sort:

By area:

  1. 2026DART: Semantic Recoverability for Structured Tool Agentscited by 10
  2. 2026Tracing Agentic Failure from the Flow of Successcited by 8
  3. 2026TRAJDEBUG: Tracing Error Lifecycle to Identify Critical Failures in Long-Horizon Agent Trajectoriescited by 8
  4. 2026Beyond Individual Intelligence: Surveying Collaboration, Failure Attribution, and Self-Evolution in LLM-based Multi-Agent Systemscited by 6
  5. 2026VerifyMAS: Hypothesis Verification for Failure Attribution in LLM Multi-Agent Systemscited by 5
  6. 2026Evolve as a Team: Collaborative Self-Evolution for LLM-based Multi-Agent Systemscited by 4
  7. 2026Long-Horizon Agent Trajectory Attribution: A Unified Benchmark and Fine-Grained Annotation Frameworkcited by 4
  8. 2026Graph Engineering in the Era of LLM Agents: From Individual Intelligence to System Intelligencecited by 3
  9. 2026AgentGrad: Intervention-guided Prompt Optimization for Multi Agent Systemscited by 3
  10. 2026LongRCA Bench: Root-Cause Localization in Long-Horizon Agent Trajectoriescited by 2
  11. 2026CockpitHAT: Dependency-Graph-Driven Hierarchical Attribution for Embodied Multi-Agent Cockpitscited by 1
  12. 2026ASCon: A Direction-Aware Reciprocal Agent--Step Contextualization Model for Failure Attribution in Multi-Agent Systemscited by 1
12 more
  1. 2026When Agentic Executions Fail: Detecting and Localizing Runtime Faults from Telemetrycited by 1
  2. 2026Adaptive Influence Graphs for Failure Attribution in Multi-Agent Systemscited by 1
  3. 2026Detect Before You Attribute: Cascade Failure Attribution for Multi-Agent Systemscited by 1
  4. 2026Agent System Operations: Categorization, Challenges, and Future Directionscited by 0
  5. 2026AgentPProf: Semantic Profiler for Long Horizon AI Agentscited by 0
  6. 2026Contextual Causality with Large Language Models: A Surveycited by 0
  7. 2026CausalLoss-Fin: Attributing Financial-Agent Loss to Decisions and Infrastructure Faultscited by 0
  8. 2026DAAF: From Failure Localization to Editable System Assets in LLM Agentscited by 0
  9. 2026Auditing Agent Actions through Query-Conditioned Attributioncited by 0
  10. 2026Structural Baselines Matter for Agent Failure Detection and Attributioncited by 0
  11. 2026Auditing Failure Attribution: Label Predictability and Judge Behaviorcited by 0
  12. 2026Reading Right, Acting Wrong: Diagnosing Decision Failures of System One Models in Agent Loopscited by 0

Similar, not cited

Close in wording, no citation either way

  1. 2026Seeing the Whole Elephant: A Benchmark for Failure Attribution in LLM-based Multi-Agent Systems
  2. 2025Which Agent Causes Task Failures and When? On Automated Failure Attribution of LLM Multi-Agent Systems
  3. 2026StepFinder: A Temporal Semantic Framework for Failure Attribution in Multi-Agent Systems
  4. 2026Error-Propagation Modeling for Failure Attribution in LLM-Based Multi-Agent Systems
  5. 2025Who is Introducing the Failure? Automatically Attributing Failures of Multi-Agent Systems via Spectrum Analysis
  6. 2026MASPrism: Lightweight Failure Attribution for Multi-Agent Systems Using Prefill-Stage Signals
  7. 2026From Flat Logs to Causal Graphs: Hierarchical Failure Attribution for LLM-based Multi-Agent Systems
  8. 2026Who&When Pro: Can LLMs Really Attribute Failures in AI Agents?