{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2025:UIZHQD2Y7S46SLOVU2XTCKST6I","short_pith_number":"pith:UIZHQD2Y","schema_version":"1.0","canonical_sha256":"a232780f58fcb9e92dd5a6af312a53f23ec7581d2f880ef18cc033e63f8fc184","source":{"kind":"arxiv","id":"2506.14704","version":1},"attestation_state":"computed","paper":{"title":"Capacity Matters: a Proof-of-Concept for Transformer Memorization on Real-World Data","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.CL","authors_text":"Aki H\\\"arm\\\"a, Anton Changalidis","submitted_at":"2025-06-17T16:42:54Z","abstract_excerpt":"This paper studies how the model architecture and data configurations influence the empirical memorization capacity of generative transformers. The models are trained using synthetic text datasets derived from the Systematized Nomenclature of Medicine (SNOMED) knowledge graph: triplets, representing static connections, and sequences, simulating complex relation patterns. The results show that embedding size is the primary determinant of learning speed and capacity, while additional layers provide limited benefits and may hinder performance on simpler datasets. Activation functions play a cruci"},"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":"2506.14704","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2025-06-17T16:42:54Z","cross_cats_sorted":[],"title_canon_sha256":"c639d640355385473a4b7fff75ac58520a6d169d2f45c56fbd2637e878fc0b73","abstract_canon_sha256":"c8bd666d50eeec9244d409363083e16bce38f9be7a1379a63ad1fc3c6a7e605a"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T11:23:11.526404Z","signature_b64":"IpJb6Vpfl+dFcEMlHZGCVxSI8sniWWm2gUVPEDGw4JTKT3VejRsb2RMURpM8x5DxI4F9TvzADniDk2pcdss7Bw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"a232780f58fcb9e92dd5a6af312a53f23ec7581d2f880ef18cc033e63f8fc184","last_reissued_at":"2026-07-05T11:23:11.525882Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T11:23:11.525882Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Capacity Matters: a Proof-of-Concept for Transformer Memorization on Real-World Data","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.CL","authors_text":"Aki H\\\"arm\\\"a, Anton Changalidis","submitted_at":"2025-06-17T16:42:54Z","abstract_excerpt":"This paper studies how the model architecture and data configurations influence the empirical memorization capacity of generative transformers. The models are trained using synthetic text datasets derived from the Systematized Nomenclature of Medicine (SNOMED) knowledge graph: triplets, representing static connections, and sequences, simulating complex relation patterns. The results show that embedding size is the primary determinant of learning speed and capacity, while additional layers provide limited benefits and may hinder performance on simpler datasets. Activation functions play a cruci"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2506.14704","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/2506.14704/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":"2506.14704","created_at":"2026-07-05T11:23:11.525943+00:00"},{"alias_kind":"arxiv_version","alias_value":"2506.14704v1","created_at":"2026-07-05T11:23:11.525943+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2506.14704","created_at":"2026-07-05T11:23:11.525943+00:00"},{"alias_kind":"pith_short_12","alias_value":"UIZHQD2Y7S46","created_at":"2026-07-05T11:23:11.525943+00:00"},{"alias_kind":"pith_short_16","alias_value":"UIZHQD2Y7S46SLOV","created_at":"2026-07-05T11:23:11.525943+00:00"},{"alias_kind":"pith_short_8","alias_value":"UIZHQD2Y","created_at":"2026-07-05T11:23:11.525943+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/UIZHQD2Y7S46SLOVU2XTCKST6I","json":"https://pith.science/pith/UIZHQD2Y7S46SLOVU2XTCKST6I.json","graph_json":"https://pith.science/api/pith-number/UIZHQD2Y7S46SLOVU2XTCKST6I/graph.json","events_json":"https://pith.science/api/pith-number/UIZHQD2Y7S46SLOVU2XTCKST6I/events.json","paper":"https://pith.science/paper/UIZHQD2Y"},"agent_actions":{"view_html":"https://pith.science/pith/UIZHQD2Y7S46SLOVU2XTCKST6I","download_json":"https://pith.science/pith/UIZHQD2Y7S46SLOVU2XTCKST6I.json","view_paper":"https://pith.science/paper/UIZHQD2Y","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2506.14704&json=true","fetch_graph":"https://pith.science/api/pith-number/UIZHQD2Y7S46SLOVU2XTCKST6I/graph.json","fetch_events":"https://pith.science/api/pith-number/UIZHQD2Y7S46SLOVU2XTCKST6I/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/UIZHQD2Y7S46SLOVU2XTCKST6I/action/timestamp_anchor","attest_storage":"https://pith.science/pith/UIZHQD2Y7S46SLOVU2XTCKST6I/action/storage_attestation","attest_author":"https://pith.science/pith/UIZHQD2Y7S46SLOVU2XTCKST6I/action/author_attestation","sign_citation":"https://pith.science/pith/UIZHQD2Y7S46SLOVU2XTCKST6I/action/citation_signature","submit_replication":"https://pith.science/pith/UIZHQD2Y7S46SLOVU2XTCKST6I/action/replication_record"}},"created_at":"2026-07-05T11:23:11.525943+00:00","updated_at":"2026-07-05T11:23:11.525943+00:00"}