{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2026:MEZ6FP32BARR6DWXVN6C3LS56P","short_pith_number":"pith:MEZ6FP32","schema_version":"1.0","canonical_sha256":"6133e2bf7a08231f0ed7ab7c2dae5df3e8021b045561c0223e14f46c19d690fe","source":{"kind":"arxiv","id":"2608.11034","version":1},"attestation_state":"computed","paper":{"title":"SCOUT: Symmetric Consensus Outlier Detection for Failure Localization in LLM Pre-Training","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.LG"],"primary_cat":"cs.DC","authors_text":"Zhuang Wang","submitted_at":"2026-08-11T15:12:14Z","abstract_excerpt":"In LLM pre-training, synchronization propagates rank-local stalls, slowdowns, and numerical errors into job-wide symptoms, obscuring their origin. Existing diagnosis often relies on in-process monitors that cannot report after the trainer blocks or terminates, or on post-mortem logs that preserve only synchronized symptoms; offline health tests lose the workload and operating conditions that triggered the failure. We present SCOUT, a unified runtime failure-localization framework built on one design principle: identify outliers through strict-majority consensus among equivalent replicas. SCOUT"},"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":"2608.11034","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.DC","submitted_at":"2026-08-11T15:12:14Z","cross_cats_sorted":["cs.LG"],"title_canon_sha256":"a25b76398783028880c40c5b64489fb7b9b78d18dc08ba28c7cfb288a97579e0","abstract_canon_sha256":"c42ef3abaa56bd91fb5c97a31dba9fb011103f369f5af0bc1874d399c8b612a0"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-08-12T01:24:28.309647Z","signature_b64":"MBq/4sA4JvVEyAxuoyLHJkl/UPLh+goWdrIPTCz8pgwZrfIlcOCN3FKbSEuyaYf6y0sNdlJGs30kTdUUFmGuBg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"6133e2bf7a08231f0ed7ab7c2dae5df3e8021b045561c0223e14f46c19d690fe","last_reissued_at":"2026-08-12T01:24:28.307562Z","signature_status":"signed_v1","first_computed_at":"2026-08-12T01:24:28.307562Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"SCOUT: Symmetric Consensus Outlier Detection for Failure Localization in LLM Pre-Training","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.LG"],"primary_cat":"cs.DC","authors_text":"Zhuang Wang","submitted_at":"2026-08-11T15:12:14Z","abstract_excerpt":"In LLM pre-training, synchronization propagates rank-local stalls, slowdowns, and numerical errors into job-wide symptoms, obscuring their origin. Existing diagnosis often relies on in-process monitors that cannot report after the trainer blocks or terminates, or on post-mortem logs that preserve only synchronized symptoms; offline health tests lose the workload and operating conditions that triggered the failure. We present SCOUT, a unified runtime failure-localization framework built on one design principle: identify outliers through strict-majority consensus among equivalent replicas. SCOUT"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2608.11034","kind":"arxiv","version":1},"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/2608.11034/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":"2608.11034","created_at":"2026-08-12T01:24:28.311466+00:00"},{"alias_kind":"arxiv_version","alias_value":"2608.11034v1","created_at":"2026-08-12T01:24:28.311466+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2608.11034","created_at":"2026-08-12T01:24:28.311466+00:00"},{"alias_kind":"pith_short_12","alias_value":"MEZ6FP32BARR","created_at":"2026-08-12T01:24:28.311466+00:00"},{"alias_kind":"pith_short_16","alias_value":"MEZ6FP32BARR6DWX","created_at":"2026-08-12T01:24:28.311466+00:00"},{"alias_kind":"pith_short_8","alias_value":"MEZ6FP32","created_at":"2026-08-12T01:24:28.311466+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":0,"internal_anchor_count":0,"sample":[]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/MEZ6FP32BARR6DWXVN6C3LS56P","json":"https://pith.science/pith/MEZ6FP32BARR6DWXVN6C3LS56P.json","graph_json":"https://pith.science/api/pith-number/MEZ6FP32BARR6DWXVN6C3LS56P/graph.json","events_json":"https://pith.science/api/pith-number/MEZ6FP32BARR6DWXVN6C3LS56P/events.json","paper":"https://pith.science/paper/MEZ6FP32"},"agent_actions":{"view_html":"https://pith.science/pith/MEZ6FP32BARR6DWXVN6C3LS56P","download_json":"https://pith.science/pith/MEZ6FP32BARR6DWXVN6C3LS56P.json","view_paper":"https://pith.science/paper/MEZ6FP32","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2608.11034&json=true","fetch_graph":"https://pith.science/api/pith-number/MEZ6FP32BARR6DWXVN6C3LS56P/graph.json","fetch_events":"https://pith.science/api/pith-number/MEZ6FP32BARR6DWXVN6C3LS56P/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/MEZ6FP32BARR6DWXVN6C3LS56P/action/timestamp_anchor","attest_storage":"https://pith.science/pith/MEZ6FP32BARR6DWXVN6C3LS56P/action/storage_attestation","attest_author":"https://pith.science/pith/MEZ6FP32BARR6DWXVN6C3LS56P/action/author_attestation","sign_citation":"https://pith.science/pith/MEZ6FP32BARR6DWXVN6C3LS56P/action/citation_signature","submit_replication":"https://pith.science/pith/MEZ6FP32BARR6DWXVN6C3LS56P/action/replication_record"}},"created_at":"2026-08-12T01:24:28.311466+00:00","updated_at":"2026-08-12T01:24:28.311466+00:00"}