{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2026:CAINK6HYBCFEJRS32TTCYWF4LI","short_pith_number":"pith:CAINK6HY","schema_version":"1.0","canonical_sha256":"1010d578f8088a44c65bd4e62c58bc5a24da69af02ccb8f23c0239e624b0b2a8","source":{"kind":"arxiv","id":"2605.25492","version":1},"attestation_state":"computed","paper":{"title":"SafetyRepro: Configuration-Conditional Rank Instability on Alignment Benchmarks","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.LG","authors_text":"Yanhang Li, Zexin Zhuang, Zhichao Fan","submitted_at":"2026-05-25T06:54:44Z","abstract_excerpt":"Pairwise model comparisons drawn from foundation-model benchmarks (\"A is safer than B\") are read as quantitative verdicts but hinge on harness choices benchmark papers under-specify. We close one theory-benchmark loop on this primitive: a finite-envelope proposition tying a measurable pairwise-disagreement rate to whether the strict ordering admits a configuration-pair reversal, paired with a commit-stamped evaluation protocol that operationalises it on widely cited alignment benchmarks. On every benchmark we test, configuration choice alone can flip the pairwise verdict; the proposition isola"},"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":"2605.25492","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2026-05-25T06:54:44Z","cross_cats_sorted":[],"title_canon_sha256":"5d9701fd056328884f9c791094c5039c10f917e7d8afebc201b72c1ad18b2530","abstract_canon_sha256":"6c4b96290cf113036d2370beb0684c1429ffa526c3c47d0c85d6f17d0b391c36"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-05-26T02:04:39.097824Z","signature_b64":"HUuEFNXicstUu6rGuqcQWKocp0vv9Kv+05rTL6ejOOwOLwLEPdmI+aYTKknqQkUYfAoZekXcaJ+uLLS1I8cWDg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"1010d578f8088a44c65bd4e62c58bc5a24da69af02ccb8f23c0239e624b0b2a8","last_reissued_at":"2026-05-26T02:04:39.096893Z","signature_status":"signed_v1","first_computed_at":"2026-05-26T02:04:39.096893Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"SafetyRepro: Configuration-Conditional Rank Instability on Alignment Benchmarks","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.LG","authors_text":"Yanhang Li, Zexin Zhuang, Zhichao Fan","submitted_at":"2026-05-25T06:54:44Z","abstract_excerpt":"Pairwise model comparisons drawn from foundation-model benchmarks (\"A is safer than B\") are read as quantitative verdicts but hinge on harness choices benchmark papers under-specify. We close one theory-benchmark loop on this primitive: a finite-envelope proposition tying a measurable pairwise-disagreement rate to whether the strict ordering admits a configuration-pair reversal, paired with a commit-stamped evaluation protocol that operationalises it on widely cited alignment benchmarks. On every benchmark we test, configuration choice alone can flip the pairwise verdict; the proposition isola"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2605.25492","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/2605.25492/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":"2605.25492","created_at":"2026-05-26T02:04:39.097045+00:00"},{"alias_kind":"arxiv_version","alias_value":"2605.25492v1","created_at":"2026-05-26T02:04:39.097045+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2605.25492","created_at":"2026-05-26T02:04:39.097045+00:00"},{"alias_kind":"pith_short_12","alias_value":"CAINK6HYBCFE","created_at":"2026-05-26T02:04:39.097045+00:00"},{"alias_kind":"pith_short_16","alias_value":"CAINK6HYBCFEJRS3","created_at":"2026-05-26T02:04:39.097045+00:00"},{"alias_kind":"pith_short_8","alias_value":"CAINK6HY","created_at":"2026-05-26T02:04:39.097045+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/CAINK6HYBCFEJRS32TTCYWF4LI","json":"https://pith.science/pith/CAINK6HYBCFEJRS32TTCYWF4LI.json","graph_json":"https://pith.science/api/pith-number/CAINK6HYBCFEJRS32TTCYWF4LI/graph.json","events_json":"https://pith.science/api/pith-number/CAINK6HYBCFEJRS32TTCYWF4LI/events.json","paper":"https://pith.science/paper/CAINK6HY"},"agent_actions":{"view_html":"https://pith.science/pith/CAINK6HYBCFEJRS32TTCYWF4LI","download_json":"https://pith.science/pith/CAINK6HYBCFEJRS32TTCYWF4LI.json","view_paper":"https://pith.science/paper/CAINK6HY","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2605.25492&json=true","fetch_graph":"https://pith.science/api/pith-number/CAINK6HYBCFEJRS32TTCYWF4LI/graph.json","fetch_events":"https://pith.science/api/pith-number/CAINK6HYBCFEJRS32TTCYWF4LI/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/CAINK6HYBCFEJRS32TTCYWF4LI/action/timestamp_anchor","attest_storage":"https://pith.science/pith/CAINK6HYBCFEJRS32TTCYWF4LI/action/storage_attestation","attest_author":"https://pith.science/pith/CAINK6HYBCFEJRS32TTCYWF4LI/action/author_attestation","sign_citation":"https://pith.science/pith/CAINK6HYBCFEJRS32TTCYWF4LI/action/citation_signature","submit_replication":"https://pith.science/pith/CAINK6HYBCFEJRS32TTCYWF4LI/action/replication_record"}},"created_at":"2026-05-26T02:04:39.097045+00:00","updated_at":"2026-05-26T02:04:39.097045+00:00"}