{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2025:IYRTCUACHANBEZDD4YGUJ4QOYV","short_pith_number":"pith:IYRTCUAC","schema_version":"1.0","canonical_sha256":"4623315002381a126463e60d44f20ec578459975af457b31b1add27b40285e6a","source":{"kind":"arxiv","id":"2510.15982","version":3},"attestation_state":"computed","paper":{"title":"AMiD: Knowledge Distillation for LLMs with $\\alpha$-mixture Assistant Distribution","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.LG","authors_text":"Byeonghu Na, Donghyeok Shin, Il-Chul Moon, Suhyeon Jo, Yeongmin Kim","submitted_at":"2025-10-13T11:05:27Z","abstract_excerpt":"Autoregressive large language models (LLMs) have achieved remarkable improvement across many tasks but incur high computational and memory costs. Knowledge distillation (KD) mitigates this issue by transferring knowledge from a large teacher to a smaller student through distributional alignment. Previous studies have proposed various discrepancy metrics, but the capacity gap and training instability caused by near-zero probabilities, stemming from the high-dimensional output of LLMs, remain fundamental limitations. To overcome these challenges, several approaches implicitly or explicitly incor"},"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":"2510.15982","kind":"arxiv","version":3},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2025-10-13T11:05:27Z","cross_cats_sorted":["cs.AI"],"title_canon_sha256":"dc0369ece20542e578fcfe2fb9da0193e34dd04018c0f8a46d316f6e0045918a","abstract_canon_sha256":"1c2f985b75a8f762bbe4da488eb146314ebd17512241e71a6f8e43e3a14e7fed"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-05-17T23:39:17.293270Z","signature_b64":"ptFM9df9ubBx2f8+CiKMeIYZerq1UVotvz/drIQPEVQR+M1JBa3l/SgxKJy+D5tSXHl0e3Hlicb4Wp7865N0AA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"4623315002381a126463e60d44f20ec578459975af457b31b1add27b40285e6a","last_reissued_at":"2026-05-17T23:39:17.292557Z","signature_status":"signed_v1","first_computed_at":"2026-05-17T23:39:17.292557Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"AMiD: Knowledge Distillation for LLMs with $\\alpha$-mixture Assistant Distribution","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.LG","authors_text":"Byeonghu Na, Donghyeok Shin, Il-Chul Moon, Suhyeon Jo, Yeongmin Kim","submitted_at":"2025-10-13T11:05:27Z","abstract_excerpt":"Autoregressive large language models (LLMs) have achieved remarkable improvement across many tasks but incur high computational and memory costs. Knowledge distillation (KD) mitigates this issue by transferring knowledge from a large teacher to a smaller student through distributional alignment. Previous studies have proposed various discrepancy metrics, but the capacity gap and training instability caused by near-zero probabilities, stemming from the high-dimensional output of LLMs, remain fundamental limitations. To overcome these challenges, several approaches implicitly or explicitly incor"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2510.15982","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":""},"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":"2510.15982","created_at":"2026-05-17T23:39:17.292656+00:00"},{"alias_kind":"arxiv_version","alias_value":"2510.15982v3","created_at":"2026-05-17T23:39:17.292656+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2510.15982","created_at":"2026-05-17T23:39:17.292656+00:00"},{"alias_kind":"pith_short_12","alias_value":"IYRTCUACHANB","created_at":"2026-05-18T12:33:37.589309+00:00"},{"alias_kind":"pith_short_16","alias_value":"IYRTCUACHANBEZDD","created_at":"2026-05-18T12:33:37.589309+00:00"},{"alias_kind":"pith_short_8","alias_value":"IYRTCUAC","created_at":"2026-05-18T12:33:37.589309+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/IYRTCUACHANBEZDD4YGUJ4QOYV","json":"https://pith.science/pith/IYRTCUACHANBEZDD4YGUJ4QOYV.json","graph_json":"https://pith.science/api/pith-number/IYRTCUACHANBEZDD4YGUJ4QOYV/graph.json","events_json":"https://pith.science/api/pith-number/IYRTCUACHANBEZDD4YGUJ4QOYV/events.json","paper":"https://pith.science/paper/IYRTCUAC"},"agent_actions":{"view_html":"https://pith.science/pith/IYRTCUACHANBEZDD4YGUJ4QOYV","download_json":"https://pith.science/pith/IYRTCUACHANBEZDD4YGUJ4QOYV.json","view_paper":"https://pith.science/paper/IYRTCUAC","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2510.15982&json=true","fetch_graph":"https://pith.science/api/pith-number/IYRTCUACHANBEZDD4YGUJ4QOYV/graph.json","fetch_events":"https://pith.science/api/pith-number/IYRTCUACHANBEZDD4YGUJ4QOYV/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/IYRTCUACHANBEZDD4YGUJ4QOYV/action/timestamp_anchor","attest_storage":"https://pith.science/pith/IYRTCUACHANBEZDD4YGUJ4QOYV/action/storage_attestation","attest_author":"https://pith.science/pith/IYRTCUACHANBEZDD4YGUJ4QOYV/action/author_attestation","sign_citation":"https://pith.science/pith/IYRTCUACHANBEZDD4YGUJ4QOYV/action/citation_signature","submit_replication":"https://pith.science/pith/IYRTCUACHANBEZDD4YGUJ4QOYV/action/replication_record"}},"created_at":"2026-05-17T23:39:17.292656+00:00","updated_at":"2026-05-17T23:39:17.292656+00:00"}