{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2024:HMC2RF6BUIZQ7BMCNLRCMBR55G","short_pith_number":"pith:HMC2RF6B","schema_version":"1.0","canonical_sha256":"3b05a897c1a2330f85826ae226063de9bc6d33d1034e458782f8580cc462093d","source":{"kind":"arxiv","id":"2406.15480","version":2},"attestation_state":"computed","paper":{"title":"On Giant's Shoulders: Effortless Weak to Strong by Dynamic Logits Fusion","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI","cs.LG"],"primary_cat":"cs.CL","authors_text":"Chenghao Fan, Dangyang Chen, Jie Tian, Wei Wei, Xiaoye Qu, Yu Cheng, Zhenyi Lu","submitted_at":"2024-06-17T03:07:41Z","abstract_excerpt":"Efficient fine-tuning of large language models for task-specific applications is imperative, yet the vast number of parameters in these models makes their training increasingly challenging. Despite numerous proposals for effective methods, a substantial memory overhead remains for gradient computations during updates. \\thm{Can we fine-tune a series of task-specific small models and transfer their knowledge directly to a much larger model without additional training?} In this paper, we explore weak-to-strong specialization using logit arithmetic, facilitating a direct answer to this question. E"},"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":"2406.15480","kind":"arxiv","version":2},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2024-06-17T03:07:41Z","cross_cats_sorted":["cs.AI","cs.LG"],"title_canon_sha256":"be54e2e1982600870fa7e5ac17e7c247e6c451e960f24a431f95bf72b061a133","abstract_canon_sha256":"75690764db42530bc3352a62abbc7a05a1f21bfef272b63cdc006515e303f8dd"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T09:20:00.091911Z","signature_b64":"OU7zJ1zyO9RfimkeipGF7/+uj3+b19B4bVya20IVgBHqwIXcRJIh/gE0zic3RzjQuS32s45mr084/uUfnSs1Dw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"3b05a897c1a2330f85826ae226063de9bc6d33d1034e458782f8580cc462093d","last_reissued_at":"2026-07-05T09:20:00.091480Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T09:20:00.091480Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"On Giant's Shoulders: Effortless Weak to Strong by Dynamic Logits Fusion","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI","cs.LG"],"primary_cat":"cs.CL","authors_text":"Chenghao Fan, Dangyang Chen, Jie Tian, Wei Wei, Xiaoye Qu, Yu Cheng, Zhenyi Lu","submitted_at":"2024-06-17T03:07:41Z","abstract_excerpt":"Efficient fine-tuning of large language models for task-specific applications is imperative, yet the vast number of parameters in these models makes their training increasingly challenging. Despite numerous proposals for effective methods, a substantial memory overhead remains for gradient computations during updates. \\thm{Can we fine-tune a series of task-specific small models and transfer their knowledge directly to a much larger model without additional training?} In this paper, we explore weak-to-strong specialization using logit arithmetic, facilitating a direct answer to this question. E"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2406.15480","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/2406.15480/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":"2406.15480","created_at":"2026-07-05T09:20:00.091538+00:00"},{"alias_kind":"arxiv_version","alias_value":"2406.15480v2","created_at":"2026-07-05T09:20:00.091538+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2406.15480","created_at":"2026-07-05T09:20:00.091538+00:00"},{"alias_kind":"pith_short_12","alias_value":"HMC2RF6BUIZQ","created_at":"2026-07-05T09:20:00.091538+00:00"},{"alias_kind":"pith_short_16","alias_value":"HMC2RF6BUIZQ7BMC","created_at":"2026-07-05T09:20:00.091538+00:00"},{"alias_kind":"pith_short_8","alias_value":"HMC2RF6B","created_at":"2026-07-05T09:20:00.091538+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/HMC2RF6BUIZQ7BMCNLRCMBR55G","json":"https://pith.science/pith/HMC2RF6BUIZQ7BMCNLRCMBR55G.json","graph_json":"https://pith.science/api/pith-number/HMC2RF6BUIZQ7BMCNLRCMBR55G/graph.json","events_json":"https://pith.science/api/pith-number/HMC2RF6BUIZQ7BMCNLRCMBR55G/events.json","paper":"https://pith.science/paper/HMC2RF6B"},"agent_actions":{"view_html":"https://pith.science/pith/HMC2RF6BUIZQ7BMCNLRCMBR55G","download_json":"https://pith.science/pith/HMC2RF6BUIZQ7BMCNLRCMBR55G.json","view_paper":"https://pith.science/paper/HMC2RF6B","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2406.15480&json=true","fetch_graph":"https://pith.science/api/pith-number/HMC2RF6BUIZQ7BMCNLRCMBR55G/graph.json","fetch_events":"https://pith.science/api/pith-number/HMC2RF6BUIZQ7BMCNLRCMBR55G/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/HMC2RF6BUIZQ7BMCNLRCMBR55G/action/timestamp_anchor","attest_storage":"https://pith.science/pith/HMC2RF6BUIZQ7BMCNLRCMBR55G/action/storage_attestation","attest_author":"https://pith.science/pith/HMC2RF6BUIZQ7BMCNLRCMBR55G/action/author_attestation","sign_citation":"https://pith.science/pith/HMC2RF6BUIZQ7BMCNLRCMBR55G/action/citation_signature","submit_replication":"https://pith.science/pith/HMC2RF6BUIZQ7BMCNLRCMBR55G/action/replication_record"}},"created_at":"2026-07-05T09:20:00.091538+00:00","updated_at":"2026-07-05T09:20:00.091538+00:00"}