{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2025:QEGI2KCQD32DNW2JKBOPW6IIIW","short_pith_number":"pith:QEGI2KCQ","schema_version":"1.0","canonical_sha256":"810c8d28501ef436db49505cfb790845886c2238aa178668062ed51f903d35d6","source":{"kind":"arxiv","id":"2505.00979","version":3},"attestation_state":"computed","paper":{"title":"Synthesize-on-Graph: Knowledgeable Synthetic Data Generation for Continue Pre-training of Large Language Models","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.CL","authors_text":"Cehao Yang, Chengjin Xu, Jian Guo, Liyu Zhang, Shengjie Ma, Xuhui Jiang","submitted_at":"2025-05-02T03:40:39Z","abstract_excerpt":"Large Language Models (LLMs) have achieved remarkable success but remain data-inefficient, especially when learning from small, specialized corpora with limited and proprietary data. Existing synthetic data generation methods for continue pre-training focus on intra-document content and overlook cross-document knowledge associations, limiting content diversity and depth. We propose Synthetic-on-Graph (SoG), a synthetic data generation framework that incorporates cross-document knowledge associations for efficient corpus expansion. SoG constructs a context graph by extracting entities and conce"},"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":"2505.00979","kind":"arxiv","version":3},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2025-05-02T03:40:39Z","cross_cats_sorted":["cs.AI"],"title_canon_sha256":"6b02286cc4f8de9af0a54154d471372673f0029f02f67813def6a3182ee6d1c3","abstract_canon_sha256":"e9961a3c22c74980f4b6528f938f91f7a1bf45b91b2af570f63d937193076ee0"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T12:11:17.573011Z","signature_b64":"00XkbhyUmTWmulSf+g2egQAffqSiNgd/7YW/hmB7g6pb/Qm1WjIovXWdutxe9+AcuV3XUr5Y55O6uL3qxoy7Bw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"810c8d28501ef436db49505cfb790845886c2238aa178668062ed51f903d35d6","last_reissued_at":"2026-07-05T12:11:17.572300Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T12:11:17.572300Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Synthesize-on-Graph: Knowledgeable Synthetic Data Generation for Continue Pre-training of Large Language Models","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.CL","authors_text":"Cehao Yang, Chengjin Xu, Jian Guo, Liyu Zhang, Shengjie Ma, Xuhui Jiang","submitted_at":"2025-05-02T03:40:39Z","abstract_excerpt":"Large Language Models (LLMs) have achieved remarkable success but remain data-inefficient, especially when learning from small, specialized corpora with limited and proprietary data. Existing synthetic data generation methods for continue pre-training focus on intra-document content and overlook cross-document knowledge associations, limiting content diversity and depth. We propose Synthetic-on-Graph (SoG), a synthetic data generation framework that incorporates cross-document knowledge associations for efficient corpus expansion. SoG constructs a context graph by extracting entities and conce"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2505.00979","kind":"arxiv","version":3},"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/2505.00979/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":"2505.00979","created_at":"2026-07-05T12:11:17.572408+00:00"},{"alias_kind":"arxiv_version","alias_value":"2505.00979v3","created_at":"2026-07-05T12:11:17.572408+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2505.00979","created_at":"2026-07-05T12:11:17.572408+00:00"},{"alias_kind":"pith_short_12","alias_value":"QEGI2KCQD32D","created_at":"2026-07-05T12:11:17.572408+00:00"},{"alias_kind":"pith_short_16","alias_value":"QEGI2KCQD32DNW2J","created_at":"2026-07-05T12:11:17.572408+00:00"},{"alias_kind":"pith_short_8","alias_value":"QEGI2KCQ","created_at":"2026-07-05T12:11:17.572408+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":1,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2605.19394","citing_title":"EmbGen: Teaching with Reassembled Corpora","ref_index":16,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/QEGI2KCQD32DNW2JKBOPW6IIIW","json":"https://pith.science/pith/QEGI2KCQD32DNW2JKBOPW6IIIW.json","graph_json":"https://pith.science/api/pith-number/QEGI2KCQD32DNW2JKBOPW6IIIW/graph.json","events_json":"https://pith.science/api/pith-number/QEGI2KCQD32DNW2JKBOPW6IIIW/events.json","paper":"https://pith.science/paper/QEGI2KCQ"},"agent_actions":{"view_html":"https://pith.science/pith/QEGI2KCQD32DNW2JKBOPW6IIIW","download_json":"https://pith.science/pith/QEGI2KCQD32DNW2JKBOPW6IIIW.json","view_paper":"https://pith.science/paper/QEGI2KCQ","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2505.00979&json=true","fetch_graph":"https://pith.science/api/pith-number/QEGI2KCQD32DNW2JKBOPW6IIIW/graph.json","fetch_events":"https://pith.science/api/pith-number/QEGI2KCQD32DNW2JKBOPW6IIIW/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/QEGI2KCQD32DNW2JKBOPW6IIIW/action/timestamp_anchor","attest_storage":"https://pith.science/pith/QEGI2KCQD32DNW2JKBOPW6IIIW/action/storage_attestation","attest_author":"https://pith.science/pith/QEGI2KCQD32DNW2JKBOPW6IIIW/action/author_attestation","sign_citation":"https://pith.science/pith/QEGI2KCQD32DNW2JKBOPW6IIIW/action/citation_signature","submit_replication":"https://pith.science/pith/QEGI2KCQD32DNW2JKBOPW6IIIW/action/replication_record"}},"created_at":"2026-07-05T12:11:17.572408+00:00","updated_at":"2026-07-05T12:11:17.572408+00:00"}