{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2026:PCQSNZBOFN5NUBJYMR3ILX2Y7W","short_pith_number":"pith:PCQSNZBO","schema_version":"1.0","canonical_sha256":"78a126e42e2b7ada0538647685df58fdbe34ee142e5487c5e428211ccdd0681a","source":{"kind":"arxiv","id":"2608.04519","version":1},"attestation_state":"computed","paper":{"title":"Leak-Resistant Unlearning: A New Benchmark for Evaluating Multi-Hop Reasoning Consistency and Recovery Robustness","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.CL"],"primary_cat":"cs.AI","authors_text":"Cheng Hong, Han Qiu, Haoting Qian, Qingjie Zhang, Zhicong Huang","submitted_at":"2026-08-05T06:53:20Z","abstract_excerpt":"Benchmarking machine unlearning methods is critical to understand whether sensitive knowledge is removed from large language models (LLMs) or not. Current unlearning benchmarks include mainly single-hop questions and a narrow set of multi-hop questions. Although effective, they still face two challenges. (1) Knowledge is not isolated, whereby diverse multi-hop reasoning paths can potentially induce knowledge leakage than normal queries. (2) Unlearning may be fragile: unlearned knowledge can be partially recovered through recovery attacks such as lightweight post-unlearning adaptation, making s"},"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":"2608.04519","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.AI","submitted_at":"2026-08-05T06:53:20Z","cross_cats_sorted":["cs.CL"],"title_canon_sha256":"af2f3004ded60c74381020bbaa1ef4194ec35c51d347a05b79563c22637b9da7","abstract_canon_sha256":"54c169176e3019d33644f973f833385c1a3b80422d516506c0cb3fdbb8f0a36f"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-08-06T01:35:46.287433Z","signature_b64":"yIrgJabVCbZgJq5vXR69Af3h/aCMwy/XQmW8uZApPyBFsXcDhtGPEuZnSKqh2oFs7f2GWnCVYq0pw16dZXapAg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"78a126e42e2b7ada0538647685df58fdbe34ee142e5487c5e428211ccdd0681a","last_reissued_at":"2026-08-06T01:35:46.285888Z","signature_status":"signed_v1","first_computed_at":"2026-08-06T01:35:46.285888Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Leak-Resistant Unlearning: A New Benchmark for Evaluating Multi-Hop Reasoning Consistency and Recovery Robustness","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.CL"],"primary_cat":"cs.AI","authors_text":"Cheng Hong, Han Qiu, Haoting Qian, Qingjie Zhang, Zhicong Huang","submitted_at":"2026-08-05T06:53:20Z","abstract_excerpt":"Benchmarking machine unlearning methods is critical to understand whether sensitive knowledge is removed from large language models (LLMs) or not. Current unlearning benchmarks include mainly single-hop questions and a narrow set of multi-hop questions. Although effective, they still face two challenges. (1) Knowledge is not isolated, whereby diverse multi-hop reasoning paths can potentially induce knowledge leakage than normal queries. (2) Unlearning may be fragile: unlearned knowledge can be partially recovered through recovery attacks such as lightweight post-unlearning adaptation, making s"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2608.04519","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/2608.04519/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":"2608.04519","created_at":"2026-08-06T01:35:46.287748+00:00"},{"alias_kind":"arxiv_version","alias_value":"2608.04519v1","created_at":"2026-08-06T01:35:46.287748+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2608.04519","created_at":"2026-08-06T01:35:46.287748+00:00"},{"alias_kind":"pith_short_12","alias_value":"PCQSNZBOFN5N","created_at":"2026-08-06T01:35:46.287748+00:00"},{"alias_kind":"pith_short_16","alias_value":"PCQSNZBOFN5NUBJY","created_at":"2026-08-06T01:35:46.287748+00:00"},{"alias_kind":"pith_short_8","alias_value":"PCQSNZBO","created_at":"2026-08-06T01:35:46.287748+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/PCQSNZBOFN5NUBJYMR3ILX2Y7W","json":"https://pith.science/pith/PCQSNZBOFN5NUBJYMR3ILX2Y7W.json","graph_json":"https://pith.science/api/pith-number/PCQSNZBOFN5NUBJYMR3ILX2Y7W/graph.json","events_json":"https://pith.science/api/pith-number/PCQSNZBOFN5NUBJYMR3ILX2Y7W/events.json","paper":"https://pith.science/paper/PCQSNZBO"},"agent_actions":{"view_html":"https://pith.science/pith/PCQSNZBOFN5NUBJYMR3ILX2Y7W","download_json":"https://pith.science/pith/PCQSNZBOFN5NUBJYMR3ILX2Y7W.json","view_paper":"https://pith.science/paper/PCQSNZBO","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2608.04519&json=true","fetch_graph":"https://pith.science/api/pith-number/PCQSNZBOFN5NUBJYMR3ILX2Y7W/graph.json","fetch_events":"https://pith.science/api/pith-number/PCQSNZBOFN5NUBJYMR3ILX2Y7W/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/PCQSNZBOFN5NUBJYMR3ILX2Y7W/action/timestamp_anchor","attest_storage":"https://pith.science/pith/PCQSNZBOFN5NUBJYMR3ILX2Y7W/action/storage_attestation","attest_author":"https://pith.science/pith/PCQSNZBOFN5NUBJYMR3ILX2Y7W/action/author_attestation","sign_citation":"https://pith.science/pith/PCQSNZBOFN5NUBJYMR3ILX2Y7W/action/citation_signature","submit_replication":"https://pith.science/pith/PCQSNZBOFN5NUBJYMR3ILX2Y7W/action/replication_record"}},"created_at":"2026-08-06T01:35:46.287748+00:00","updated_at":"2026-08-06T01:35:46.287748+00:00"}