{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2026:HENXVC6I4MFEBY6RC2JJTMCXVO","short_pith_number":"pith:HENXVC6I","schema_version":"1.0","canonical_sha256":"391b7a8bc8e30a40e3d1169299b057abb6529067d855b4059081c9309e3fc76b","source":{"kind":"arxiv","id":"2605.30393","version":1},"attestation_state":"computed","paper":{"title":"NumLeak: Public Numeric Benchmarks as Latent Labels in Foundation Models","license":"http://creativecommons.org/licenses/by-nc-sa/4.0/","headline":"","cross_cats":["cs.AI","cs.CR"],"primary_cat":"cs.LG","authors_text":"Anany Kotawala","submitted_at":"2026-05-28T12:52:49Z","abstract_excerpt":"Public numeric benchmarks appear in pretraining, so an evaluation that conditions on a date may be measuring memorized recall rather than out-of-sample skill. We introduce NumLeak, a measurement framework that combines API-boundary probes on production models with a white-box controlled validation on an open causal LM. Top-tier frontier LLMs recall the Fama-French market excess return at 3-seed pooled Pearson r=0.97-0.99 while staying within 0.15 within-25bps on the five sibling factors; comparable fidelity appears on U.S. unemployment, CPI inflation, and NOAA temperature. On a recent-release "},"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":"2605.30393","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by-nc-sa/4.0/","primary_cat":"cs.LG","submitted_at":"2026-05-28T12:52:49Z","cross_cats_sorted":["cs.AI","cs.CR"],"title_canon_sha256":"e37854d1ccfe0e799a706ac12dfb128fea56024ca24438bc4ee421d3faf37777","abstract_canon_sha256":"9bbec6b00fd4d120754ef39c38d4f6bad0f7039dd8a8f0e0f15df36ad298bf61"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-06-01T00:02:08.909432Z","signature_b64":"idz6nVj/o8WncdO8giMobJgaWvQxJczyjKwjX8DPUoBdRRA19G+SBZLeJHCscPUZuXED8fIW1785OFJ1SjHNBg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"391b7a8bc8e30a40e3d1169299b057abb6529067d855b4059081c9309e3fc76b","last_reissued_at":"2026-06-01T00:02:08.908716Z","signature_status":"signed_v1","first_computed_at":"2026-06-01T00:02:08.908716Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"NumLeak: Public Numeric Benchmarks as Latent Labels in Foundation Models","license":"http://creativecommons.org/licenses/by-nc-sa/4.0/","headline":"","cross_cats":["cs.AI","cs.CR"],"primary_cat":"cs.LG","authors_text":"Anany Kotawala","submitted_at":"2026-05-28T12:52:49Z","abstract_excerpt":"Public numeric benchmarks appear in pretraining, so an evaluation that conditions on a date may be measuring memorized recall rather than out-of-sample skill. We introduce NumLeak, a measurement framework that combines API-boundary probes on production models with a white-box controlled validation on an open causal LM. Top-tier frontier LLMs recall the Fama-French market excess return at 3-seed pooled Pearson r=0.97-0.99 while staying within 0.15 within-25bps on the five sibling factors; comparable fidelity appears on U.S. unemployment, CPI inflation, and NOAA temperature. On a recent-release "},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2605.30393","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/2605.30393/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":"2605.30393","created_at":"2026-06-01T00:02:08.908808+00:00"},{"alias_kind":"arxiv_version","alias_value":"2605.30393v1","created_at":"2026-06-01T00:02:08.908808+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2605.30393","created_at":"2026-06-01T00:02:08.908808+00:00"},{"alias_kind":"pith_short_12","alias_value":"HENXVC6I4MFE","created_at":"2026-06-01T00:02:08.908808+00:00"},{"alias_kind":"pith_short_16","alias_value":"HENXVC6I4MFEBY6R","created_at":"2026-06-01T00:02:08.908808+00:00"},{"alias_kind":"pith_short_8","alias_value":"HENXVC6I","created_at":"2026-06-01T00:02:08.908808+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/HENXVC6I4MFEBY6RC2JJTMCXVO","json":"https://pith.science/pith/HENXVC6I4MFEBY6RC2JJTMCXVO.json","graph_json":"https://pith.science/api/pith-number/HENXVC6I4MFEBY6RC2JJTMCXVO/graph.json","events_json":"https://pith.science/api/pith-number/HENXVC6I4MFEBY6RC2JJTMCXVO/events.json","paper":"https://pith.science/paper/HENXVC6I"},"agent_actions":{"view_html":"https://pith.science/pith/HENXVC6I4MFEBY6RC2JJTMCXVO","download_json":"https://pith.science/pith/HENXVC6I4MFEBY6RC2JJTMCXVO.json","view_paper":"https://pith.science/paper/HENXVC6I","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2605.30393&json=true","fetch_graph":"https://pith.science/api/pith-number/HENXVC6I4MFEBY6RC2JJTMCXVO/graph.json","fetch_events":"https://pith.science/api/pith-number/HENXVC6I4MFEBY6RC2JJTMCXVO/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/HENXVC6I4MFEBY6RC2JJTMCXVO/action/timestamp_anchor","attest_storage":"https://pith.science/pith/HENXVC6I4MFEBY6RC2JJTMCXVO/action/storage_attestation","attest_author":"https://pith.science/pith/HENXVC6I4MFEBY6RC2JJTMCXVO/action/author_attestation","sign_citation":"https://pith.science/pith/HENXVC6I4MFEBY6RC2JJTMCXVO/action/citation_signature","submit_replication":"https://pith.science/pith/HENXVC6I4MFEBY6RC2JJTMCXVO/action/replication_record"}},"created_at":"2026-06-01T00:02:08.908808+00:00","updated_at":"2026-06-01T00:02:08.908808+00:00"}