{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2026:YHEMZCTAE6YZTFQAFAY2KFCJ4U","short_pith_number":"pith:YHEMZCTA","schema_version":"1.0","canonical_sha256":"c1c8cc8a6027b19996002831a51449e503f461d3bae3951bcf32be82162dfe46","source":{"kind":"arxiv","id":"2603.00191","version":4},"attestation_state":"computed","paper":{"title":"Task-Driven Subspace Decomposition for Knowledge Sharing and Isolation in LoRA-based Continual Learning","license":"http://creativecommons.org/licenses/by/4.0/","headline":"LoDA decomposes LoRA updates into shared general and task-specific subspaces via energy-based objectives to enable knowledge transfer without catastrophic forgetting.","cross_cats":["cs.CV"],"primary_cat":"cs.LG","authors_text":"De Cheng, Huaijie Wang, Lingfeng He, Nannan Wang, Xinbo Gao, Xi Yang","submitted_at":"2026-02-27T02:31:00Z","abstract_excerpt":"Continual Learning (CL) requires models to sequentially adapt to new tasks without forgetting old knowledge. Recently, Low-Rank Adaptation (LoRA), a representative Parameter-Efficient Fine-Tuning (PEFT) method, has gained increasing attention in CL. Several LoRA-based CL methods reduce interference across tasks by separating their update spaces, typically building the new space from the estimated null space of past tasks. However, they (i) overlook task-shared directions, which suppresses knowledge transfer, and (ii) fail to capture truly effective task-specific directions since these ``null b"},"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":true},"canonical_record":{"source":{"id":"2603.00191","kind":"arxiv","version":4},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2026-02-27T02:31:00Z","cross_cats_sorted":["cs.CV"],"title_canon_sha256":"1af17290de1aca09b5df77ed7c7430c707b128208f440cc46ca9745b644b2781","abstract_canon_sha256":"b8808c3d788d985ec7101547ff37cbe6e98595941672cf5a8b6b34f8312ea4f3"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-05-26T01:03:26.855497Z","signature_b64":"i4Spwesa+f7q7Gs0GWYTGnjXredqd4FfahoNLvUXUxYkPNzugXDAXmUhDLz7MhRdL/UJadErObVTU39f8oRaCw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"c1c8cc8a6027b19996002831a51449e503f461d3bae3951bcf32be82162dfe46","last_reissued_at":"2026-05-26T01:03:26.854634Z","signature_status":"signed_v1","first_computed_at":"2026-05-26T01:03:26.854634Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Task-Driven Subspace Decomposition for Knowledge Sharing and Isolation in LoRA-based Continual Learning","license":"http://creativecommons.org/licenses/by/4.0/","headline":"LoDA decomposes LoRA updates into shared general and task-specific subspaces via energy-based objectives to enable knowledge transfer without catastrophic forgetting.","cross_cats":["cs.CV"],"primary_cat":"cs.LG","authors_text":"De Cheng, Huaijie Wang, Lingfeng He, Nannan Wang, Xinbo Gao, Xi Yang","submitted_at":"2026-02-27T02:31:00Z","abstract_excerpt":"Continual Learning (CL) requires models to sequentially adapt to new tasks without forgetting old knowledge. Recently, Low-Rank Adaptation (LoRA), a representative Parameter-Efficient Fine-Tuning (PEFT) method, has gained increasing attention in CL. Several LoRA-based CL methods reduce interference across tasks by separating their update spaces, typically building the new space from the estimated null space of past tasks. However, they (i) overlook task-shared directions, which suppresses knowledge transfer, and (ii) fail to capture truly effective task-specific directions since these ``null b"},"claims":{"count":4,"items":[{"kind":"strongest_claim","text":"LoDA outperforms existing CL methods by performing task-driven decomposition to build general and truly task-specific LoRA subspaces, fixing down-projections and using GAO plus closed-form recalibration for the general update.","source":"verdict.strongest_claim","status":"machine_extracted","claim_id":"C1","attestation":"unclaimed"},{"kind":"weakest_assumption","text":"That solving the two energy-based objectives will reliably produce effective shared and task-specific directions, and that the null-space limitations of prior methods are the primary bottleneck rather than other factors such as optimization dynamics or data distribution shifts.","source":"verdict.weakest_assumption","status":"machine_extracted","claim_id":"C2","attestation":"unclaimed"},{"kind":"one_line_summary","text":"LoDA introduces task-driven subspace decomposition in LoRA for continual learning to separate knowledge-sharing and isolation directions via energy objectives and closed-form recalibration, outperforming prior null-space methods.","source":"verdict.one_line_summary","status":"machine_extracted","claim_id":"C3","attestation":"unclaimed"},{"kind":"headline","text":"LoDA decomposes LoRA updates into shared general and task-specific subspaces via energy-based objectives to enable knowledge transfer without catastrophic forgetting.","source":"verdict.pith_extraction.headline","status":"machine_extracted","claim_id":"C4","attestation":"unclaimed"}],"snapshot_sha256":"7c8ed58ac5108b75aff7fa0846d4f5367d92ca76946dfe310a1aae805db71049"},"source":{"id":"2603.00191","kind":"arxiv","version":4},"verdict":{"id":"d06e7625-f9bd-41a7-b0f4-9733442cc3e6","model_set":{"reader":"grok-4.3"},"created_at":"2026-05-15T18:57:44.672935Z","strongest_claim":"LoDA outperforms existing CL methods by performing task-driven decomposition to build general and truly task-specific LoRA subspaces, fixing down-projections and using GAO plus closed-form recalibration for the general update.","one_line_summary":"LoDA introduces task-driven subspace decomposition in LoRA for continual learning to separate knowledge-sharing and isolation directions via energy objectives and closed-form recalibration, outperforming prior null-space methods.","pipeline_version":"pith-pipeline@v0.9.0","weakest_assumption":"That solving the two energy-based objectives will reliably produce effective shared and task-specific directions, and that the null-space limitations of prior methods are the primary bottleneck rather than other factors such as optimization dynamics or data distribution shifts.","pith_extraction_headline":"LoDA decomposes LoRA updates into shared general and task-specific subspaces via energy-based objectives to enable knowledge transfer without catastrophic forgetting."},"integrity":{"clean":true,"summary":{"advisory":0,"critical":0,"by_detector":{},"informational":0},"endpoint":"/pith/2603.00191/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":2,"snapshot_sha256":"9f2b9ada3c4306b0400aef4cfb5c269e4a13701c96cc272741c1871248e1619d"},"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":"2603.00191","created_at":"2026-05-26T01:03:26.854766+00:00"},{"alias_kind":"arxiv_version","alias_value":"2603.00191v4","created_at":"2026-05-26T01:03:26.854766+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2603.00191","created_at":"2026-05-26T01:03:26.854766+00:00"},{"alias_kind":"pith_short_12","alias_value":"YHEMZCTAE6YZ","created_at":"2026-05-26T01:03:26.854766+00:00"},{"alias_kind":"pith_short_16","alias_value":"YHEMZCTAE6YZTFQA","created_at":"2026-05-26T01:03:26.854766+00:00"},{"alias_kind":"pith_short_8","alias_value":"YHEMZCTA","created_at":"2026-05-26T01:03:26.854766+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":0,"internal_anchor_count":0,"sample":[]},"formal_canon":{"evidence_count":2,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/YHEMZCTAE6YZTFQAFAY2KFCJ4U","json":"https://pith.science/pith/YHEMZCTAE6YZTFQAFAY2KFCJ4U.json","graph_json":"https://pith.science/api/pith-number/YHEMZCTAE6YZTFQAFAY2KFCJ4U/graph.json","events_json":"https://pith.science/api/pith-number/YHEMZCTAE6YZTFQAFAY2KFCJ4U/events.json","paper":"https://pith.science/paper/YHEMZCTA"},"agent_actions":{"view_html":"https://pith.science/pith/YHEMZCTAE6YZTFQAFAY2KFCJ4U","download_json":"https://pith.science/pith/YHEMZCTAE6YZTFQAFAY2KFCJ4U.json","view_paper":"https://pith.science/paper/YHEMZCTA","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2603.00191&json=true","fetch_graph":"https://pith.science/api/pith-number/YHEMZCTAE6YZTFQAFAY2KFCJ4U/graph.json","fetch_events":"https://pith.science/api/pith-number/YHEMZCTAE6YZTFQAFAY2KFCJ4U/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/YHEMZCTAE6YZTFQAFAY2KFCJ4U/action/timestamp_anchor","attest_storage":"https://pith.science/pith/YHEMZCTAE6YZTFQAFAY2KFCJ4U/action/storage_attestation","attest_author":"https://pith.science/pith/YHEMZCTAE6YZTFQAFAY2KFCJ4U/action/author_attestation","sign_citation":"https://pith.science/pith/YHEMZCTAE6YZTFQAFAY2KFCJ4U/action/citation_signature","submit_replication":"https://pith.science/pith/YHEMZCTAE6YZTFQAFAY2KFCJ4U/action/replication_record"}},"created_at":"2026-05-26T01:03:26.854766+00:00","updated_at":"2026-05-26T01:03:26.854766+00:00"}