{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2026:H74STC72BK7WODWTLIT534S42V","short_pith_number":"pith:H74STC72","schema_version":"1.0","canonical_sha256":"3ff9298bfa0abf670ed35a27ddf25cd54bffa6fc20388fc4dd0d72d73c746f75","source":{"kind":"arxiv","id":"2606.29159","version":1},"attestation_state":"computed","paper":{"title":"Pooled Leaderboards Hide System-Specific Winners: A Reporting-Protocol Audit of Offline Root-Cause Analysis Benchmarks","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.AI","authors_text":"Lining Hu, Ting Liu, Yuzhuo Fu","submitted_at":"2026-06-28T02:47:31Z","abstract_excerpt":"Offline root-cause-analysis (RCA) benchmarks commonly rank methods by a single pooled top-1 accuracy across multiple subsystems, and engineers often read the pooled winner as a recommendation for their own subsystem. We audit that reading on three public RCA benchmark families -- OpenRCA, RCAEval, and PetShop -- covering 11 subsystems and 778 matched scoring units. To keep pairwise comparisons on identical cases, the main analysis retains four methods or comparators with complete coverage: BARO, a CD-1min adapter, max-$|Z|$, and per-service alert-count. All six pairwise comparisons show subsys"},"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":"2606.29159","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.AI","submitted_at":"2026-06-28T02:47:31Z","cross_cats_sorted":[],"title_canon_sha256":"4fe64508d5575b47845ce922aa816a16c59b55bab0918e5cdee6eda1c261fcdf","abstract_canon_sha256":"79f0869d57e75b77914fdfd5ca69636515dafd516cee09d703b3e5ea10bef01d"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-06-30T01:17:55.031761Z","signature_b64":"Sf/Dpb2vjrOGHUM0wTsOHeKzCW+U3JVTgiaZuuE3NP1eDGqTHFhboDWwq9i1qPnERg+KawaaKg8VPd0OnB32Ag==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"3ff9298bfa0abf670ed35a27ddf25cd54bffa6fc20388fc4dd0d72d73c746f75","last_reissued_at":"2026-06-30T01:17:55.031110Z","signature_status":"signed_v1","first_computed_at":"2026-06-30T01:17:55.031110Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Pooled Leaderboards Hide System-Specific Winners: A Reporting-Protocol Audit of Offline Root-Cause Analysis Benchmarks","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.AI","authors_text":"Lining Hu, Ting Liu, Yuzhuo Fu","submitted_at":"2026-06-28T02:47:31Z","abstract_excerpt":"Offline root-cause-analysis (RCA) benchmarks commonly rank methods by a single pooled top-1 accuracy across multiple subsystems, and engineers often read the pooled winner as a recommendation for their own subsystem. We audit that reading on three public RCA benchmark families -- OpenRCA, RCAEval, and PetShop -- covering 11 subsystems and 778 matched scoring units. To keep pairwise comparisons on identical cases, the main analysis retains four methods or comparators with complete coverage: BARO, a CD-1min adapter, max-$|Z|$, and per-service alert-count. All six pairwise comparisons show subsys"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2606.29159","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/2606.29159/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":"2606.29159","created_at":"2026-06-30T01:17:55.031178+00:00"},{"alias_kind":"arxiv_version","alias_value":"2606.29159v1","created_at":"2026-06-30T01:17:55.031178+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2606.29159","created_at":"2026-06-30T01:17:55.031178+00:00"},{"alias_kind":"pith_short_12","alias_value":"H74STC72BK7W","created_at":"2026-06-30T01:17:55.031178+00:00"},{"alias_kind":"pith_short_16","alias_value":"H74STC72BK7WODWT","created_at":"2026-06-30T01:17:55.031178+00:00"},{"alias_kind":"pith_short_8","alias_value":"H74STC72","created_at":"2026-06-30T01:17:55.031178+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/H74STC72BK7WODWTLIT534S42V","json":"https://pith.science/pith/H74STC72BK7WODWTLIT534S42V.json","graph_json":"https://pith.science/api/pith-number/H74STC72BK7WODWTLIT534S42V/graph.json","events_json":"https://pith.science/api/pith-number/H74STC72BK7WODWTLIT534S42V/events.json","paper":"https://pith.science/paper/H74STC72"},"agent_actions":{"view_html":"https://pith.science/pith/H74STC72BK7WODWTLIT534S42V","download_json":"https://pith.science/pith/H74STC72BK7WODWTLIT534S42V.json","view_paper":"https://pith.science/paper/H74STC72","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2606.29159&json=true","fetch_graph":"https://pith.science/api/pith-number/H74STC72BK7WODWTLIT534S42V/graph.json","fetch_events":"https://pith.science/api/pith-number/H74STC72BK7WODWTLIT534S42V/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/H74STC72BK7WODWTLIT534S42V/action/timestamp_anchor","attest_storage":"https://pith.science/pith/H74STC72BK7WODWTLIT534S42V/action/storage_attestation","attest_author":"https://pith.science/pith/H74STC72BK7WODWTLIT534S42V/action/author_attestation","sign_citation":"https://pith.science/pith/H74STC72BK7WODWTLIT534S42V/action/citation_signature","submit_replication":"https://pith.science/pith/H74STC72BK7WODWTLIT534S42V/action/replication_record"}},"created_at":"2026-06-30T01:17:55.031178+00:00","updated_at":"2026-06-30T01:17:55.031178+00:00"}