{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2024:DXSP5NVEERJHRQQZWEVKF5WFKC","short_pith_number":"pith:DXSP5NVE","schema_version":"1.0","canonical_sha256":"1de4feb6a4245278c219b12aa2f6c5508189a12c4b8b44279dba6123b7483b0d","source":{"kind":"arxiv","id":"2405.18380","version":3},"attestation_state":"computed","paper":{"title":"Outlier-weighed Layerwise Sampling for LLM Fine-tuning","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI","cs.CL"],"primary_cat":"cs.LG","authors_text":"Lu Yin, Pengxiang Li, Shiwei Liu, Xiaowei Gao","submitted_at":"2024-05-28T17:22:22Z","abstract_excerpt":"The rapid advancements in Large Language Models (LLMs) have revolutionized various natural language processing tasks. However, the substantial size of LLMs presents significant challenges in training or fine-tuning. While parameter-efficient approaches such as low-rank adaptation (LoRA) have gained popularity, they often compromise performance compared to full-rank fine-tuning. In this paper, we propose Outlier-weighed Layerwise Sampling (OWS), a new memory-efficient fine-tuning approach, inspired by the layerwise outlier distribution of LLMs. Unlike LoRA, which adds extra adapters to all laye"},"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":"2405.18380","kind":"arxiv","version":3},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2024-05-28T17:22:22Z","cross_cats_sorted":["cs.AI","cs.CL"],"title_canon_sha256":"d0a02d09f67b69d1350036aec011ca1386a84dddd226ccf84f57c55e932fe3d6","abstract_canon_sha256":"f961797764d2159cb317f605409b76ee6d3d8365417ad2387fa584332d3241da"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T11:18:12.423080Z","signature_b64":"RMFLjdql2HWjd+HKRl4yrMEobwwUfyhF4wiXIOt1OTu3By1fl0pGHoDj1yrsI6StthNNbdMdqc9qtbcJxKATDQ==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"1de4feb6a4245278c219b12aa2f6c5508189a12c4b8b44279dba6123b7483b0d","last_reissued_at":"2026-07-05T11:18:12.422628Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T11:18:12.422628Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Outlier-weighed Layerwise Sampling for LLM Fine-tuning","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI","cs.CL"],"primary_cat":"cs.LG","authors_text":"Lu Yin, Pengxiang Li, Shiwei Liu, Xiaowei Gao","submitted_at":"2024-05-28T17:22:22Z","abstract_excerpt":"The rapid advancements in Large Language Models (LLMs) have revolutionized various natural language processing tasks. However, the substantial size of LLMs presents significant challenges in training or fine-tuning. While parameter-efficient approaches such as low-rank adaptation (LoRA) have gained popularity, they often compromise performance compared to full-rank fine-tuning. In this paper, we propose Outlier-weighed Layerwise Sampling (OWS), a new memory-efficient fine-tuning approach, inspired by the layerwise outlier distribution of LLMs. Unlike LoRA, which adds extra adapters to all laye"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2405.18380","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":""},"integrity":{"clean":true,"summary":{"advisory":0,"critical":0,"by_detector":{},"informational":0},"endpoint":"/pith/2405.18380/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":"2405.18380","created_at":"2026-07-05T11:18:12.422685+00:00"},{"alias_kind":"arxiv_version","alias_value":"2405.18380v3","created_at":"2026-07-05T11:18:12.422685+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2405.18380","created_at":"2026-07-05T11:18:12.422685+00:00"},{"alias_kind":"pith_short_12","alias_value":"DXSP5NVEERJH","created_at":"2026-07-05T11:18:12.422685+00:00"},{"alias_kind":"pith_short_16","alias_value":"DXSP5NVEERJHRQQZ","created_at":"2026-07-05T11:18:12.422685+00:00"},{"alias_kind":"pith_short_8","alias_value":"DXSP5NVE","created_at":"2026-07-05T11:18:12.422685+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":1,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2412.00069","citing_title":"Condense, Don't Just Prune: Enhancing Efficiency and Performance in MoE Layer Pruning","ref_index":21,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/DXSP5NVEERJHRQQZWEVKF5WFKC","json":"https://pith.science/pith/DXSP5NVEERJHRQQZWEVKF5WFKC.json","graph_json":"https://pith.science/api/pith-number/DXSP5NVEERJHRQQZWEVKF5WFKC/graph.json","events_json":"https://pith.science/api/pith-number/DXSP5NVEERJHRQQZWEVKF5WFKC/events.json","paper":"https://pith.science/paper/DXSP5NVE"},"agent_actions":{"view_html":"https://pith.science/pith/DXSP5NVEERJHRQQZWEVKF5WFKC","download_json":"https://pith.science/pith/DXSP5NVEERJHRQQZWEVKF5WFKC.json","view_paper":"https://pith.science/paper/DXSP5NVE","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2405.18380&json=true","fetch_graph":"https://pith.science/api/pith-number/DXSP5NVEERJHRQQZWEVKF5WFKC/graph.json","fetch_events":"https://pith.science/api/pith-number/DXSP5NVEERJHRQQZWEVKF5WFKC/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/DXSP5NVEERJHRQQZWEVKF5WFKC/action/timestamp_anchor","attest_storage":"https://pith.science/pith/DXSP5NVEERJHRQQZWEVKF5WFKC/action/storage_attestation","attest_author":"https://pith.science/pith/DXSP5NVEERJHRQQZWEVKF5WFKC/action/author_attestation","sign_citation":"https://pith.science/pith/DXSP5NVEERJHRQQZWEVKF5WFKC/action/citation_signature","submit_replication":"https://pith.science/pith/DXSP5NVEERJHRQQZWEVKF5WFKC/action/replication_record"}},"created_at":"2026-07-05T11:18:12.422685+00:00","updated_at":"2026-07-05T11:18:12.422685+00:00"}