{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2026:N3TZSJ5DVO6HNXUOTEGPNZVBMA","short_pith_number":"pith:N3TZSJ5D","schema_version":"1.0","canonical_sha256":"6ee79927a3abbc76de8e990cf6e6a160193e465e42c973f425f41bd5a5b7c9a8","source":{"kind":"arxiv","id":"2605.16379","version":1},"attestation_state":"computed","paper":{"title":"An Information-Theoretic Criterion for Efficient Data Synthesis","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI","cs.IT","math.IT"],"primary_cat":"cs.LG","authors_text":"Hanyu Li, Xiaotie Deng, Zhengqi Sun","submitted_at":"2026-05-11T01:27:59Z","abstract_excerpt":"Synthetic data becomes crucial for large language model training, but its effectiveness is highly inconsistent. We provide an information-theoretic account of this inconsistency: synthetic data improves a model only when the generation-training loop is information-open, i.e., shaped by external signals (verifiers, environments, or rubrics) that inject task-relevant information beyond the model's current distribution. When the loop is information-closed (relying on the model's own outputs without such signals), the data processing inequality ensures that task-relevant information can only decre"},"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":"2605.16379","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2026-05-11T01:27:59Z","cross_cats_sorted":["cs.AI","cs.IT","math.IT"],"title_canon_sha256":"c557e5f898197b49bc6ebe4586a6e4a29d2e95b736db4aba223bb3092906a9d2","abstract_canon_sha256":"fe69b79b04134bf4781fc9420e72fc89008cde5d9be091f6f33e880471da289a"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-05-20T00:02:19.249532Z","signature_b64":"y2sBAor6HcjpffUsrl/ESLPQXq6tznTir9MRkgb/JKwOWfKf79FmwOoY0eGU2xB6MuQZvRyHLAj71STstIdGCQ==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"6ee79927a3abbc76de8e990cf6e6a160193e465e42c973f425f41bd5a5b7c9a8","last_reissued_at":"2026-05-20T00:02:19.248846Z","signature_status":"signed_v1","first_computed_at":"2026-05-20T00:02:19.248846Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"An Information-Theoretic Criterion for Efficient Data Synthesis","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI","cs.IT","math.IT"],"primary_cat":"cs.LG","authors_text":"Hanyu Li, Xiaotie Deng, Zhengqi Sun","submitted_at":"2026-05-11T01:27:59Z","abstract_excerpt":"Synthetic data becomes crucial for large language model training, but its effectiveness is highly inconsistent. We provide an information-theoretic account of this inconsistency: synthetic data improves a model only when the generation-training loop is information-open, i.e., shaped by external signals (verifiers, environments, or rubrics) that inject task-relevant information beyond the model's current distribution. When the loop is information-closed (relying on the model's own outputs without such signals), the data processing inequality ensures that task-relevant information can only decre"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2605.16379","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/2605.16379/integrity.json","findings":[],"available":true,"detectors_run":[{"name":"ai_meta_artifact","ran_at":"2026-05-19T19:34:36.615102Z","status":"skipped","version":"1.0.0","findings_count":0}],"snapshot_sha256":"a233500c8b3697ab26c30d7b67bd98eb06f9c7a41a458d98a666a2e3445e01f1"},"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":"2605.16379","created_at":"2026-05-20T00:02:19.248957+00:00"},{"alias_kind":"arxiv_version","alias_value":"2605.16379v1","created_at":"2026-05-20T00:02:19.248957+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2605.16379","created_at":"2026-05-20T00:02:19.248957+00:00"},{"alias_kind":"pith_short_12","alias_value":"N3TZSJ5DVO6H","created_at":"2026-05-20T00:02:19.248957+00:00"},{"alias_kind":"pith_short_16","alias_value":"N3TZSJ5DVO6HNXUO","created_at":"2026-05-20T00:02:19.248957+00:00"},{"alias_kind":"pith_short_8","alias_value":"N3TZSJ5D","created_at":"2026-05-20T00:02:19.248957+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/N3TZSJ5DVO6HNXUOTEGPNZVBMA","json":"https://pith.science/pith/N3TZSJ5DVO6HNXUOTEGPNZVBMA.json","graph_json":"https://pith.science/api/pith-number/N3TZSJ5DVO6HNXUOTEGPNZVBMA/graph.json","events_json":"https://pith.science/api/pith-number/N3TZSJ5DVO6HNXUOTEGPNZVBMA/events.json","paper":"https://pith.science/paper/N3TZSJ5D"},"agent_actions":{"view_html":"https://pith.science/pith/N3TZSJ5DVO6HNXUOTEGPNZVBMA","download_json":"https://pith.science/pith/N3TZSJ5DVO6HNXUOTEGPNZVBMA.json","view_paper":"https://pith.science/paper/N3TZSJ5D","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2605.16379&json=true","fetch_graph":"https://pith.science/api/pith-number/N3TZSJ5DVO6HNXUOTEGPNZVBMA/graph.json","fetch_events":"https://pith.science/api/pith-number/N3TZSJ5DVO6HNXUOTEGPNZVBMA/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/N3TZSJ5DVO6HNXUOTEGPNZVBMA/action/timestamp_anchor","attest_storage":"https://pith.science/pith/N3TZSJ5DVO6HNXUOTEGPNZVBMA/action/storage_attestation","attest_author":"https://pith.science/pith/N3TZSJ5DVO6HNXUOTEGPNZVBMA/action/author_attestation","sign_citation":"https://pith.science/pith/N3TZSJ5DVO6HNXUOTEGPNZVBMA/action/citation_signature","submit_replication":"https://pith.science/pith/N3TZSJ5DVO6HNXUOTEGPNZVBMA/action/replication_record"}},"created_at":"2026-05-20T00:02:19.248957+00:00","updated_at":"2026-05-20T00:02:19.248957+00:00"}