{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2024:YAPEDO6DX7JSSH2Y22E4664LWX","short_pith_number":"pith:YAPEDO6D","schema_version":"1.0","canonical_sha256":"c01e41bbc3bfd3291f58d689cf7b8bb5e9a62620ba1f8800f4f81688ec020656","source":{"kind":"arxiv","id":"2412.19616","version":2},"attestation_state":"computed","paper":{"title":"Gradient Weight-normalized Low-rank Projection for Efficient LLM Training","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.LG","authors_text":"Evangelos Kanoulas, Hongyi Zhu, Jia-Hong Huang, Stevan Rudinac, Yixian Shen","submitted_at":"2024-12-27T12:23:39Z","abstract_excerpt":"Large Language Models (LLMs) have shown remarkable performance across various tasks, but the escalating demands on computational resources pose significant challenges, particularly in the extensive utilization of full fine-tuning for downstream tasks. To address this, parameter-efficient fine-tuning (PEFT) methods have been developed, but they often underperform compared to full fine-tuning and struggle with memory efficiency. In this work, we introduce Gradient Weight-Normalized Low-Rank Projection (GradNormLoRP), a novel approach that enhances both parameter and memory efficiency while maint"},"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":"2412.19616","kind":"arxiv","version":2},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2024-12-27T12:23:39Z","cross_cats_sorted":["cs.AI"],"title_canon_sha256":"655f3e88f2959156d47474dce8a067362cadbd661c1e001cf78fbeed7ff42c3f","abstract_canon_sha256":"6cbed8de485e75d00fb7aaaebbc2587be0c4f7a1e246e714864b6982ded8b973"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T09:56:58.458730Z","signature_b64":"gZcxTIcYdgF/ylC3jcnsN5BrwQwaRZdZxgZR3LWtcouC1AdkxuQtt8e5jFzO6CKAj2h+VsHENWDqWsX+S1SGDw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"c01e41bbc3bfd3291f58d689cf7b8bb5e9a62620ba1f8800f4f81688ec020656","last_reissued_at":"2026-07-05T09:56:58.458273Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T09:56:58.458273Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Gradient Weight-normalized Low-rank Projection for Efficient LLM Training","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.LG","authors_text":"Evangelos Kanoulas, Hongyi Zhu, Jia-Hong Huang, Stevan Rudinac, Yixian Shen","submitted_at":"2024-12-27T12:23:39Z","abstract_excerpt":"Large Language Models (LLMs) have shown remarkable performance across various tasks, but the escalating demands on computational resources pose significant challenges, particularly in the extensive utilization of full fine-tuning for downstream tasks. To address this, parameter-efficient fine-tuning (PEFT) methods have been developed, but they often underperform compared to full fine-tuning and struggle with memory efficiency. In this work, we introduce Gradient Weight-Normalized Low-Rank Projection (GradNormLoRP), a novel approach that enhances both parameter and memory efficiency while maint"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2412.19616","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/2412.19616/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":"2412.19616","created_at":"2026-07-05T09:56:58.458326+00:00"},{"alias_kind":"arxiv_version","alias_value":"2412.19616v2","created_at":"2026-07-05T09:56:58.458326+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2412.19616","created_at":"2026-07-05T09:56:58.458326+00:00"},{"alias_kind":"pith_short_12","alias_value":"YAPEDO6DX7JS","created_at":"2026-07-05T09:56:58.458326+00:00"},{"alias_kind":"pith_short_16","alias_value":"YAPEDO6DX7JSSH2Y","created_at":"2026-07-05T09:56:58.458326+00:00"},{"alias_kind":"pith_short_8","alias_value":"YAPEDO6D","created_at":"2026-07-05T09:56:58.458326+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":1,"internal_anchor_count":1,"sample":[{"citing_arxiv_id":"2502.05539","citing_title":"SSH: Sparse Spectrum Adaptation via Discrete Hartley Transformation","ref_index":17,"is_internal_anchor":true}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/YAPEDO6DX7JSSH2Y22E4664LWX","json":"https://pith.science/pith/YAPEDO6DX7JSSH2Y22E4664LWX.json","graph_json":"https://pith.science/api/pith-number/YAPEDO6DX7JSSH2Y22E4664LWX/graph.json","events_json":"https://pith.science/api/pith-number/YAPEDO6DX7JSSH2Y22E4664LWX/events.json","paper":"https://pith.science/paper/YAPEDO6D"},"agent_actions":{"view_html":"https://pith.science/pith/YAPEDO6DX7JSSH2Y22E4664LWX","download_json":"https://pith.science/pith/YAPEDO6DX7JSSH2Y22E4664LWX.json","view_paper":"https://pith.science/paper/YAPEDO6D","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2412.19616&json=true","fetch_graph":"https://pith.science/api/pith-number/YAPEDO6DX7JSSH2Y22E4664LWX/graph.json","fetch_events":"https://pith.science/api/pith-number/YAPEDO6DX7JSSH2Y22E4664LWX/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/YAPEDO6DX7JSSH2Y22E4664LWX/action/timestamp_anchor","attest_storage":"https://pith.science/pith/YAPEDO6DX7JSSH2Y22E4664LWX/action/storage_attestation","attest_author":"https://pith.science/pith/YAPEDO6DX7JSSH2Y22E4664LWX/action/author_attestation","sign_citation":"https://pith.science/pith/YAPEDO6DX7JSSH2Y22E4664LWX/action/citation_signature","submit_replication":"https://pith.science/pith/YAPEDO6DX7JSSH2Y22E4664LWX/action/replication_record"}},"created_at":"2026-07-05T09:56:58.458326+00:00","updated_at":"2026-07-05T09:56:58.458326+00:00"}