mlsanityData to 7 Oct 2026

POIROT: Interrogating Agents for Failure Detection in Multi-Agent Systems

arXiv 2606.02282 ↗First posted 1 Jun 2026cs.AIcited by 0

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

also in 1 more
Abstract (arXiv)

Orchestrating Large Language Models into Multi-Agent Systems (LLM-MAS) has unlocked remarkable reasoning capabilities, yet emergent failures and hallucinations that resist characterisation block their deployment in safety-critical domains -- a gap made legally untenable by emerging AI regulation. Existing evaluation paradigms share a common flaw: centralised judgment creates single points of failure and demands domain-specific expertise. Here we present POIROT, a protocol that repurposes a system's own agents as its diagnostic layer, leveraging the epistemic diversity already present in the architecture. Across evaluated settings, POIROT outperforms single-LLM evaluator baselines, with gains that scale with problem complexity (OR = 1.60, p = 0.008), agent count, and fault dimensionality, persisting under compound fault conditions. These results demonstrate that safety oversight need not be externalised: the agents executing a role carry sufficient collective intelligence to audit it. We release POIROT as an open-source library alongside BLAME, a benchmark for fault attribution in safety-critical multi-agent systems.

Key facts. No citing papers in the map yet (data through 7 Oct 2026). Placed in Reasoning for agents, planning and code, one of 17 directions in LLM reasoning. Builds on 22 of its 36 references that are in the map. First posted on arXiv 1 Jun 2026.

Builds on

22 of 36 references.

References · 22

  1. 2023Judging LLM-as-a-Judge with MT-Bench and Chatbot Arenacited by 8,393
  2. 2023Self-Refine: Iterative Refinement with Self-Feedbackcited by 3,317
  3. 2022Constitutional AI: Harmlessness from AI Feedbackcited by 2,950
  4. 2024SWE-agent: Agent-Computer Interfaces Enable Automated Software Engineeringcited by 1,628
  5. 2024A Survey on LLM-as-a-Judgecited by 1,384
  6. 2023MemoryBank: Enhancing Large Language Models with Long-Term Memorycited by 539
  7. 2023Multi-Agent Collaboration: Harnessing the Power of Intelligent LLM Agentscited by 329
  8. 2023Chain-of-Verification Reduces Hallucination in Large Language Modelscited by 317
  9. 2024Humans or LLMs as the Judge? A Study on Judgement Biasescited by 314
  10. 2024Judging the Judges: Evaluating Alignment and Vulnerabilities in LLMs-as-Judgescited by 197
  11. 2024TradingAgents: Multi-Agents LLM Financial Trading Frameworkcited by 182
  12. 2025Which Agent Causes Task Failures and When? On Automated Failure Attribution of LLM Multi-Agent Systemscited by 170
10 more
  1. 2024LLM-Based Multi-Agent Systems for Software Engineering: Literature Review, Vision and the Road Aheadcited by 155
  2. 2024LLM Multi-Agent Systems: Challenges and Open Problemscited by 119
  3. 2024A Systematic Survey and Critical Review on Evaluating Large Language Models: Challenges, Limitations, and Recommendationscited by 84
  4. 2025HaluMem: Evaluating Hallucinations in Memory Systems of Agentscited by 51
  5. 2024DocETL: Agentic Query Rewriting and Evaluation for Complex Document Processingcited by 49
  6. 2025The relationship between reasoning and performance in large language models--o3 (mini) thinks harder, not longercited by 39
  7. 2025DoVer: Intervention-Driven Auto Debugging for LLM Multi-Agent Systemscited by 23
  8. 2025LLM-Friendly Knowledge Representation for Customer Supportcited by 11
  9. 2025Failure Modes in LLM Systems: A System-Level Taxonomy for Reliable AI Applicationscited by 9
  10. 2025Evaluating LLMs' Assessment of Mixed-Context Hallucination Through the Lens of Summarizationcited by 6

Built on by

Not cited yet.

Similar, not cited

Close in wording, no citation either way

  1. 2026Towards Self-Improving Error Diagnosis in Multi-Agent Systems
  2. 2026Who Broke the System? Failure Localization in LLM-Based Multi-Agent Systems
  3. 2026MAS-FIRE: Fault Injection and Reliability Evaluation for LLM-Based Multi-Agent Systems
  4. 2026Efficient Failure Management for Multi-Agent Systems with Reasoning Trace Representation
  5. 2026AgentFixer: From Failure Detection to Fix Recommendations in LLM Agentic Systems
  6. 2025CORRECT: COndensed eRror RECognition via knowledge Transfer in multi-agent systems
  7. 2026Finding Where the Buck Stops: An Automated Failure Attribution-Based Reflection Framework for Multi-Agent Collaboration
  8. 2026VerifyMAS: Hypothesis Verification for Failure Attribution in LLM Multi-Agent Systems