{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2024:DKN4MFFD3EU6H5QK2WWSUSOMSE","short_pith_number":"pith:DKN4MFFD","schema_version":"1.0","canonical_sha256":"1a9bc614a3d929e3f60ad5ad2a49cc912c9664b803b2cc95f27efa1dcdfd68f0","source":{"kind":"arxiv","id":"2408.05541","version":2},"attestation_state":"computed","paper":{"title":"P3: A Policy-Driven, Pace-Adaptive, and Diversity-Promoted Framework for data pruning in LLM Training","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.CL","authors_text":"Huayi Wang, Jun Wang, Muning Wen, Qiuying Peng, Weinan Zhang, Xiaoyun Mo, Yingxuan Yang","submitted_at":"2024-08-10T12:44:49Z","abstract_excerpt":"In the rapidly advancing field of Large Language Models (LLMs), effectively leveraging existing datasets during fine-tuning to maximize the model's potential is of paramount importance. This paper introduces P3, an adaptive framework aimed at optimizing the task-specific fine-tuning process through iterative data pruning. P3 consists of three key components: (1) Policy-driven Difficulty Measurement, which dynamically assesses data difficulty based on the model's real-time performance, replacing static metrics with adaptable evaluations; (2) Pace-Adaptive Selection, leveraging self-paced learni"},"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":"2408.05541","kind":"arxiv","version":2},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2024-08-10T12:44:49Z","cross_cats_sorted":[],"title_canon_sha256":"34c5b478a059d762ec09bb17c7ade5cdc2d80ce5f7df8db3e0a85160b5818088","abstract_canon_sha256":"6ad111b462f366e371d55e0f823bf83eac3c055debd2ecf7dda7ad9cdd36d08d"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T09:22:21.073751Z","signature_b64":"I4pHne5plOBT6Q+kV5L7yYvA27rCOnPFbVl2Bsae7fbIRPjekRGEM2rznJeWd2rLrYSZQ5DxsrpQoz9/khBHCw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"1a9bc614a3d929e3f60ad5ad2a49cc912c9664b803b2cc95f27efa1dcdfd68f0","last_reissued_at":"2026-07-05T09:22:21.073324Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T09:22:21.073324Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"P3: A Policy-Driven, Pace-Adaptive, and Diversity-Promoted Framework for data pruning in LLM Training","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.CL","authors_text":"Huayi Wang, Jun Wang, Muning Wen, Qiuying Peng, Weinan Zhang, Xiaoyun Mo, Yingxuan Yang","submitted_at":"2024-08-10T12:44:49Z","abstract_excerpt":"In the rapidly advancing field of Large Language Models (LLMs), effectively leveraging existing datasets during fine-tuning to maximize the model's potential is of paramount importance. This paper introduces P3, an adaptive framework aimed at optimizing the task-specific fine-tuning process through iterative data pruning. P3 consists of three key components: (1) Policy-driven Difficulty Measurement, which dynamically assesses data difficulty based on the model's real-time performance, replacing static metrics with adaptable evaluations; (2) Pace-Adaptive Selection, leveraging self-paced learni"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2408.05541","kind":"arxiv","version":2},"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/2408.05541/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":"2408.05541","created_at":"2026-07-05T09:22:21.073389+00:00"},{"alias_kind":"arxiv_version","alias_value":"2408.05541v2","created_at":"2026-07-05T09:22:21.073389+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2408.05541","created_at":"2026-07-05T09:22:21.073389+00:00"},{"alias_kind":"pith_short_12","alias_value":"DKN4MFFD3EU6","created_at":"2026-07-05T09:22:21.073389+00:00"},{"alias_kind":"pith_short_16","alias_value":"DKN4MFFD3EU6H5QK","created_at":"2026-07-05T09:22:21.073389+00:00"},{"alias_kind":"pith_short_8","alias_value":"DKN4MFFD","created_at":"2026-07-05T09:22:21.073389+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/DKN4MFFD3EU6H5QK2WWSUSOMSE","json":"https://pith.science/pith/DKN4MFFD3EU6H5QK2WWSUSOMSE.json","graph_json":"https://pith.science/api/pith-number/DKN4MFFD3EU6H5QK2WWSUSOMSE/graph.json","events_json":"https://pith.science/api/pith-number/DKN4MFFD3EU6H5QK2WWSUSOMSE/events.json","paper":"https://pith.science/paper/DKN4MFFD"},"agent_actions":{"view_html":"https://pith.science/pith/DKN4MFFD3EU6H5QK2WWSUSOMSE","download_json":"https://pith.science/pith/DKN4MFFD3EU6H5QK2WWSUSOMSE.json","view_paper":"https://pith.science/paper/DKN4MFFD","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2408.05541&json=true","fetch_graph":"https://pith.science/api/pith-number/DKN4MFFD3EU6H5QK2WWSUSOMSE/graph.json","fetch_events":"https://pith.science/api/pith-number/DKN4MFFD3EU6H5QK2WWSUSOMSE/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/DKN4MFFD3EU6H5QK2WWSUSOMSE/action/timestamp_anchor","attest_storage":"https://pith.science/pith/DKN4MFFD3EU6H5QK2WWSUSOMSE/action/storage_attestation","attest_author":"https://pith.science/pith/DKN4MFFD3EU6H5QK2WWSUSOMSE/action/author_attestation","sign_citation":"https://pith.science/pith/DKN4MFFD3EU6H5QK2WWSUSOMSE/action/citation_signature","submit_replication":"https://pith.science/pith/DKN4MFFD3EU6H5QK2WWSUSOMSE/action/replication_record"}},"created_at":"2026-07-05T09:22:21.073389+00:00","updated_at":"2026-07-05T09:22:21.073389+00:00"}