{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2025:AARIOWVKUIN5WIACTODRLVIF3R","short_pith_number":"pith:AARIOWVK","schema_version":"1.0","canonical_sha256":"0022875aaaa21bdb20029b8715d505dc79fd0a65f2a5aded6504a84927bbe54a","source":{"kind":"arxiv","id":"2505.00212","version":3},"attestation_state":"computed","paper":{"title":"Which Agent Causes Task Failures and When? On Automated Failure Attribution of LLM Multi-Agent Systems","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.CL"],"primary_cat":"cs.MA","authors_text":"Beibin Li, Chi Wang, Huazheng Wang, Jiale Liu, Jieyu Zhang, Jingyang Zhang, Ming Yin, Qingyun Wu, Shaokun Zhang, Yiran Chen, Zhiguang Han","submitted_at":"2025-04-30T23:09:44Z","abstract_excerpt":"Failure attribution in LLM multi-agent systems-identifying the agent and step responsible for task failures-provides crucial clues for systems debugging but remains underexplored and labor-intensive. In this paper, we propose and formulate a new research area: automated failure attribution for LLM multi-agent systems. To support this initiative, we introduce the Who&When dataset, comprising extensive failure logs from 127 LLM multi-agent systems with fine-grained annotations linking failures to specific agents and decisive error steps. Using the Who&When, we develop and evaluate three automate"},"verification_status":{"content_addressed":true,"pith_receipt":true,"author_attested":false,"weak_author_claims":0,"strong_author_claims":0,"externally_anchored":false,"storage_verified":false,"citation_signatures":0,"replication_records":0,"graph_snapshot":true,"references_resolved":false,"formal_links_present":false},"canonical_record":{"source":{"id":"2505.00212","kind":"arxiv","version":3},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.MA","submitted_at":"2025-04-30T23:09:44Z","cross_cats_sorted":["cs.CL"],"title_canon_sha256":"9d058cf8df08564a8ed9fb7c0d0f59b9d4bb9f666b3313418d297821745555c9","abstract_canon_sha256":"3275b958f6d3fcbed39a8e96109e66c7bd817209e0db327412a798cace5d3780"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T11:14:06.747104Z","signature_b64":"5rY/1VY7pPp9nexh2/kdVHjAlsVNVT3f7UfLWt6jrvvFL5FOUaxW57THcKHYXBlY0xEkVomD1Vi1FBJrEQxzDg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"0022875aaaa21bdb20029b8715d505dc79fd0a65f2a5aded6504a84927bbe54a","last_reissued_at":"2026-07-05T11:14:06.746627Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T11:14:06.746627Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Which Agent Causes Task Failures and When? On Automated Failure Attribution of LLM Multi-Agent Systems","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.CL"],"primary_cat":"cs.MA","authors_text":"Beibin Li, Chi Wang, Huazheng Wang, Jiale Liu, Jieyu Zhang, Jingyang Zhang, Ming Yin, Qingyun Wu, Shaokun Zhang, Yiran Chen, Zhiguang Han","submitted_at":"2025-04-30T23:09:44Z","abstract_excerpt":"Failure attribution in LLM multi-agent systems-identifying the agent and step responsible for task failures-provides crucial clues for systems debugging but remains underexplored and labor-intensive. In this paper, we propose and formulate a new research area: automated failure attribution for LLM multi-agent systems. To support this initiative, we introduce the Who&When dataset, comprising extensive failure logs from 127 LLM multi-agent systems with fine-grained annotations linking failures to specific agents and decisive error steps. Using the Who&When, we develop and evaluate three automate"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2505.00212","kind":"arxiv","version":3},"verdict":{"id":null,"model_set":{},"created_at":null,"strongest_claim":"","one_line_summary":"","pipeline_version":null,"weakest_assumption":"","pith_extraction_headline":""},"integrity":{"clean":true,"summary":{"advisory":0,"critical":0,"by_detector":{},"informational":0},"endpoint":"/pith/2505.00212/integrity.json","findings":[],"available":true,"detectors_run":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938"},"references":{"count":0,"sample":[],"resolved_work":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57","internal_anchors":0},"formal_canon":{"evidence_count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"author_claims":{"count":0,"strong_count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"builder_version":"pith-number-builder-2026-05-17-v1"},"aliases":[{"alias_kind":"arxiv","alias_value":"2505.00212","created_at":"2026-07-05T11:14:06.746685+00:00"},{"alias_kind":"arxiv_version","alias_value":"2505.00212v3","created_at":"2026-07-05T11:14:06.746685+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2505.00212","created_at":"2026-07-05T11:14:06.746685+00:00"},{"alias_kind":"pith_short_12","alias_value":"AARIOWVKUIN5","created_at":"2026-07-05T11:14:06.746685+00:00"},{"alias_kind":"pith_short_16","alias_value":"AARIOWVKUIN5WIAC","created_at":"2026-07-05T11:14:06.746685+00:00"},{"alias_kind":"pith_short_8","alias_value":"AARIOWVK","created_at":"2026-07-05T11:14:06.746685+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":27,"internal_anchor_count":1,"sample":[{"citing_arxiv_id":"2607.06273","citing_title":"AgentTether: Graph-Guided Diagnosis and Runtime Intervention for Reliable LLM Agent Operation","ref_index":25,"is_internal_anchor":true},{"citing_arxiv_id":"2606.23664","citing_title":"MAS-PromptBench: When Does Prompt Optimization Improve Multi-Agent LLM Systems?","ref_index":40,"is_internal_anchor":false},{"citing_arxiv_id":"2606.22741","citing_title":"GRADE: Graph Representation of LLM Agent Dependency and Execution","ref_index":144,"is_internal_anchor":false},{"citing_arxiv_id":"2606.08275","citing_title":"Causal Agent Replay: Counterfactual Attribution for LLM-Agent Failures","ref_index":1,"is_internal_anchor":false},{"citing_arxiv_id":"2606.07889","citing_title":"Strained Coherence: A Pre-Failure Signal in Coding Agent Execution Trajectories","ref_index":7,"is_internal_anchor":false},{"citing_arxiv_id":"2606.03467","citing_title":"StepFinder: A Temporal Semantic Framework for Failure Attribution in Multi-Agent Systems","ref_index":51,"is_internal_anchor":false},{"citing_arxiv_id":"2606.02282","citing_title":"POIROT: Interrogating Agents for Failure Detection in Multi-Agent Systems","ref_index":30,"is_internal_anchor":false},{"citing_arxiv_id":"2606.27409","citing_title":"Delayed Verification Destabilizes Multi-Agent LLM Belief: Instability Thresholds and Optimal Corrector Placement","ref_index":17,"is_internal_anchor":false},{"citing_arxiv_id":"2605.21347","citing_title":"Insights Generator: Systematic Corpus-Level Trace Diagnostics for LLM Agents","ref_index":25,"is_internal_anchor":false},{"citing_arxiv_id":"2605.26302","citing_title":"Your Agents Are Aging Too: Agent Lifespan Engineering for Deployed Systems","ref_index":46,"is_internal_anchor":false},{"citing_arxiv_id":"2606.00765","citing_title":"FALAT: Tracing Failures in LLM Agent Trajectories via Dependency-Guided Search","ref_index":6,"is_internal_anchor":false},{"citing_arxiv_id":"2605.23414","citing_title":"When Planning Fails Despite Correct Execution: On Epistemic Calibration for LLM-Based Multi-Agent Systems","ref_index":139,"is_internal_anchor":false},{"citing_arxiv_id":"2605.21347","citing_title":"Insights Generator: Systematic Corpus-Level Trace Diagnostics for LLM Agents","ref_index":25,"is_internal_anchor":false},{"citing_arxiv_id":"2605.21347","citing_title":"Insights Generator: Systematic Corpus-Level Trace Diagnostics for LLM Agents","ref_index":25,"is_internal_anchor":false},{"citing_arxiv_id":"2605.15581","citing_title":"STAR: A Stage-attributed Triage and Repair framework for RCA Agents in Microservices","ref_index":29,"is_internal_anchor":false},{"citing_arxiv_id":"2605.17352","citing_title":"AMATA: Adaptive Multi-Agent Trajectory Alignment for Knowledge-Intensive Question Answering","ref_index":9,"is_internal_anchor":false},{"citing_arxiv_id":"2605.19240","citing_title":"CASPIAN: Online Detection and Attribution of Cascade Attacks in LLM Multi-Agent Systems via Cross-Channel Causal Monitoring","ref_index":44,"is_internal_anchor":false},{"citing_arxiv_id":"2605.14859","citing_title":"Do Coding Agents Understand Least-Privilege Authorization?","ref_index":68,"is_internal_anchor":false},{"citing_arxiv_id":"2510.02837","citing_title":"Beyond the Final Answer: Evaluating the Reasoning Trajectories of Tool-Augmented Agents","ref_index":15,"is_internal_anchor":false},{"citing_arxiv_id":"2603.04474","citing_title":"From Spark to Fire: Modeling and Mitigating Error Cascades in LLM-Based Multi-Agent Collaboration","ref_index":56,"is_internal_anchor":false},{"citing_arxiv_id":"2603.06847","citing_title":"Characterizing Faults in Agentic AI: A Taxonomy of Types, Symptoms, and Root Causes","ref_index":45,"is_internal_anchor":false},{"citing_arxiv_id":"2605.08715","citing_title":"AgentForesight: Online Auditing for Early Failure Prediction in Multi-Agent Systems","ref_index":65,"is_internal_anchor":false},{"citing_arxiv_id":"2605.11225","citing_title":"PIVOT: Bridging Planning and Execution in LLM Agents via Trajectory Refinement","ref_index":25,"is_internal_anchor":false},{"citing_arxiv_id":"2604.27586","citing_title":"Trace-Level Analysis of Information Contamination in Multi-Agent Systems","ref_index":38,"is_internal_anchor":false},{"citing_arxiv_id":"2605.08715","citing_title":"AgentForesight: Online Auditing for Early Failure Prediction in Multi-Agent Systems","ref_index":65,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/AARIOWVKUIN5WIACTODRLVIF3R","json":"https://pith.science/pith/AARIOWVKUIN5WIACTODRLVIF3R.json","graph_json":"https://pith.science/api/pith-number/AARIOWVKUIN5WIACTODRLVIF3R/graph.json","events_json":"https://pith.science/api/pith-number/AARIOWVKUIN5WIACTODRLVIF3R/events.json","paper":"https://pith.science/paper/AARIOWVK"},"agent_actions":{"view_html":"https://pith.science/pith/AARIOWVKUIN5WIACTODRLVIF3R","download_json":"https://pith.science/pith/AARIOWVKUIN5WIACTODRLVIF3R.json","view_paper":"https://pith.science/paper/AARIOWVK","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2505.00212&json=true","fetch_graph":"https://pith.science/api/pith-number/AARIOWVKUIN5WIACTODRLVIF3R/graph.json","fetch_events":"https://pith.science/api/pith-number/AARIOWVKUIN5WIACTODRLVIF3R/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/AARIOWVKUIN5WIACTODRLVIF3R/action/timestamp_anchor","attest_storage":"https://pith.science/pith/AARIOWVKUIN5WIACTODRLVIF3R/action/storage_attestation","attest_author":"https://pith.science/pith/AARIOWVKUIN5WIACTODRLVIF3R/action/author_attestation","sign_citation":"https://pith.science/pith/AARIOWVKUIN5WIACTODRLVIF3R/action/citation_signature","submit_replication":"https://pith.science/pith/AARIOWVKUIN5WIACTODRLVIF3R/action/replication_record"}},"created_at":"2026-07-05T11:14:06.746685+00:00","updated_at":"2026-07-05T11:14:06.746685+00:00"}