{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2024:6KXVUWWZCDV2MWGIXDZYDUG2AA","short_pith_number":"pith:6KXVUWWZ","schema_version":"1.0","canonical_sha256":"f2af5a5ad910eba658c8b8f381d0da001b7c0d809201c0be43de6172ad86c8d2","source":{"kind":"arxiv","id":"2412.15303","version":1},"attestation_state":"computed","paper":{"title":"Self-Evolution Knowledge Distillation for LLM-based Machine Translation","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.CL","authors_text":"Changtong Zan, Liang Ding, Shujian Huang, Yuncheng Song","submitted_at":"2024-12-19T12:24:15Z","abstract_excerpt":"Knowledge distillation (KD) has shown great promise in transferring knowledge from larger teacher models to smaller student models. However, existing KD strategies for large language models often minimize output distributions between student and teacher models indiscriminately for each token. This overlooks the imbalanced nature of tokens and their varying transfer difficulties. In response, we propose a distillation strategy called Self-Evolution KD. The core of this approach involves dynamically integrating teacher distribution and one-hot distribution of ground truth into the student distri"},"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":"2412.15303","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2024-12-19T12:24:15Z","cross_cats_sorted":[],"title_canon_sha256":"afcdbbde9f0445ac6038ab7c07550233c1451d751e1c1433916c274764bc10b8","abstract_canon_sha256":"7e3f13a109c389c0f929e335c1e33fd916a1bee26536d06e1d99b58026747052"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T09:52:20.948059Z","signature_b64":"pzUNeRcufL2kydNSWxmu97bKKrgUNjhuzQ6OWtOGNFeeYEKBVVKcSxVEgM/zAHTjNy16UVoF64RAsoc/Zr+NDg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"f2af5a5ad910eba658c8b8f381d0da001b7c0d809201c0be43de6172ad86c8d2","last_reissued_at":"2026-07-05T09:52:20.947587Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T09:52:20.947587Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Self-Evolution Knowledge Distillation for LLM-based Machine Translation","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.CL","authors_text":"Changtong Zan, Liang Ding, Shujian Huang, Yuncheng Song","submitted_at":"2024-12-19T12:24:15Z","abstract_excerpt":"Knowledge distillation (KD) has shown great promise in transferring knowledge from larger teacher models to smaller student models. However, existing KD strategies for large language models often minimize output distributions between student and teacher models indiscriminately for each token. This overlooks the imbalanced nature of tokens and their varying transfer difficulties. In response, we propose a distillation strategy called Self-Evolution KD. The core of this approach involves dynamically integrating teacher distribution and one-hot distribution of ground truth into the student distri"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2412.15303","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/2412.15303/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":"2412.15303","created_at":"2026-07-05T09:52:20.947650+00:00"},{"alias_kind":"arxiv_version","alias_value":"2412.15303v1","created_at":"2026-07-05T09:52:20.947650+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2412.15303","created_at":"2026-07-05T09:52:20.947650+00:00"},{"alias_kind":"pith_short_12","alias_value":"6KXVUWWZCDV2","created_at":"2026-07-05T09:52:20.947650+00:00"},{"alias_kind":"pith_short_16","alias_value":"6KXVUWWZCDV2MWGI","created_at":"2026-07-05T09:52:20.947650+00:00"},{"alias_kind":"pith_short_8","alias_value":"6KXVUWWZ","created_at":"2026-07-05T09:52:20.947650+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/6KXVUWWZCDV2MWGIXDZYDUG2AA","json":"https://pith.science/pith/6KXVUWWZCDV2MWGIXDZYDUG2AA.json","graph_json":"https://pith.science/api/pith-number/6KXVUWWZCDV2MWGIXDZYDUG2AA/graph.json","events_json":"https://pith.science/api/pith-number/6KXVUWWZCDV2MWGIXDZYDUG2AA/events.json","paper":"https://pith.science/paper/6KXVUWWZ"},"agent_actions":{"view_html":"https://pith.science/pith/6KXVUWWZCDV2MWGIXDZYDUG2AA","download_json":"https://pith.science/pith/6KXVUWWZCDV2MWGIXDZYDUG2AA.json","view_paper":"https://pith.science/paper/6KXVUWWZ","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2412.15303&json=true","fetch_graph":"https://pith.science/api/pith-number/6KXVUWWZCDV2MWGIXDZYDUG2AA/graph.json","fetch_events":"https://pith.science/api/pith-number/6KXVUWWZCDV2MWGIXDZYDUG2AA/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/6KXVUWWZCDV2MWGIXDZYDUG2AA/action/timestamp_anchor","attest_storage":"https://pith.science/pith/6KXVUWWZCDV2MWGIXDZYDUG2AA/action/storage_attestation","attest_author":"https://pith.science/pith/6KXVUWWZCDV2MWGIXDZYDUG2AA/action/author_attestation","sign_citation":"https://pith.science/pith/6KXVUWWZCDV2MWGIXDZYDUG2AA/action/citation_signature","submit_replication":"https://pith.science/pith/6KXVUWWZCDV2MWGIXDZYDUG2AA/action/replication_record"}},"created_at":"2026-07-05T09:52:20.947650+00:00","updated_at":"2026-07-05T09:52:20.947650+00:00"}