{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2026:MH2SB2Z7Q5ZTT5CKKWSOIGEONL","short_pith_number":"pith:MH2SB2Z7","schema_version":"1.0","canonical_sha256":"61f520eb3f877339f44a55a4e4188e6af9331d5df642c573eb96ed6ef752b498","source":{"kind":"arxiv","id":"2605.31164","version":1},"attestation_state":"computed","paper":{"title":"D$^3$: Dynamic Directional Graph-Constrained Data Scheduling for LLM Training","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.CL","authors_text":"Guang Zhang, Jianing Hao, Yuanjian Xu, Zhong Li","submitted_at":"2026-05-29T11:13:43Z","abstract_excerpt":"Training data plays a central role in large language models (LLMs) optimization, motivating extensive research on data scheduling strategies. Most existing approaches concentrate on adjusting the overall data distribution but neglect the underlying interactions between samples during training. However, we argue that such interactions cannot be overlooked, as real-world data samples frequently exhibit directional influences on each other, making the training order crucial. Intuitively, we can prioritize train-units with greater influence to improves learning efficiency. In this work, we propose"},"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":"2605.31164","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2026-05-29T11:13:43Z","cross_cats_sorted":["cs.AI"],"title_canon_sha256":"fb5188543ad85291868e19b98d264c72257a86ce9deec9ea14ab766acae874f4","abstract_canon_sha256":"438c9b0f380ee9c3d3babcf131068cb79b2f855c2708d6b5546d51bae1b00c24"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-06-01T01:04:01.979161Z","signature_b64":"tvNEaBXOyEztavkKyBtwurgiz0ZYLwwmlD8KnszVg4cFcaPRouigLxKBCnWJ/Q2ilAtVP3EeiG3sRFQaF9LhDg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"61f520eb3f877339f44a55a4e4188e6af9331d5df642c573eb96ed6ef752b498","last_reissued_at":"2026-06-01T01:04:01.978616Z","signature_status":"signed_v1","first_computed_at":"2026-06-01T01:04:01.978616Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"D$^3$: Dynamic Directional Graph-Constrained Data Scheduling for LLM Training","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.CL","authors_text":"Guang Zhang, Jianing Hao, Yuanjian Xu, Zhong Li","submitted_at":"2026-05-29T11:13:43Z","abstract_excerpt":"Training data plays a central role in large language models (LLMs) optimization, motivating extensive research on data scheduling strategies. Most existing approaches concentrate on adjusting the overall data distribution but neglect the underlying interactions between samples during training. However, we argue that such interactions cannot be overlooked, as real-world data samples frequently exhibit directional influences on each other, making the training order crucial. Intuitively, we can prioritize train-units with greater influence to improves learning efficiency. In this work, we propose"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2605.31164","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/2605.31164/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":"2605.31164","created_at":"2026-06-01T01:04:01.978702+00:00"},{"alias_kind":"arxiv_version","alias_value":"2605.31164v1","created_at":"2026-06-01T01:04:01.978702+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2605.31164","created_at":"2026-06-01T01:04:01.978702+00:00"},{"alias_kind":"pith_short_12","alias_value":"MH2SB2Z7Q5ZT","created_at":"2026-06-01T01:04:01.978702+00:00"},{"alias_kind":"pith_short_16","alias_value":"MH2SB2Z7Q5ZTT5CK","created_at":"2026-06-01T01:04:01.978702+00:00"},{"alias_kind":"pith_short_8","alias_value":"MH2SB2Z7","created_at":"2026-06-01T01:04:01.978702+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/MH2SB2Z7Q5ZTT5CKKWSOIGEONL","json":"https://pith.science/pith/MH2SB2Z7Q5ZTT5CKKWSOIGEONL.json","graph_json":"https://pith.science/api/pith-number/MH2SB2Z7Q5ZTT5CKKWSOIGEONL/graph.json","events_json":"https://pith.science/api/pith-number/MH2SB2Z7Q5ZTT5CKKWSOIGEONL/events.json","paper":"https://pith.science/paper/MH2SB2Z7"},"agent_actions":{"view_html":"https://pith.science/pith/MH2SB2Z7Q5ZTT5CKKWSOIGEONL","download_json":"https://pith.science/pith/MH2SB2Z7Q5ZTT5CKKWSOIGEONL.json","view_paper":"https://pith.science/paper/MH2SB2Z7","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2605.31164&json=true","fetch_graph":"https://pith.science/api/pith-number/MH2SB2Z7Q5ZTT5CKKWSOIGEONL/graph.json","fetch_events":"https://pith.science/api/pith-number/MH2SB2Z7Q5ZTT5CKKWSOIGEONL/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/MH2SB2Z7Q5ZTT5CKKWSOIGEONL/action/timestamp_anchor","attest_storage":"https://pith.science/pith/MH2SB2Z7Q5ZTT5CKKWSOIGEONL/action/storage_attestation","attest_author":"https://pith.science/pith/MH2SB2Z7Q5ZTT5CKKWSOIGEONL/action/author_attestation","sign_citation":"https://pith.science/pith/MH2SB2Z7Q5ZTT5CKKWSOIGEONL/action/citation_signature","submit_replication":"https://pith.science/pith/MH2SB2Z7Q5ZTT5CKKWSOIGEONL/action/replication_record"}},"created_at":"2026-06-01T01:04:01.978702+00:00","updated_at":"2026-06-01T01:04:01.978702+00:00"}