{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2025:HMIBALIASOGBCZVTZ3CCGI5OXM","short_pith_number":"pith:HMIBALIA","schema_version":"1.0","canonical_sha256":"3b10102d00938c1166b3cec42323aebb2c4dca28d92941bf8163f9cf4c8b7374","source":{"kind":"arxiv","id":"2502.06029","version":3},"attestation_state":"computed","paper":{"title":"DiTASK: Multi-Task Fine-Tuning with Diffeomorphic Transformations","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.CV","authors_text":"Bruno Ribeiro, Carola-Bibiane Sch\\\"onlieb, Chaim Baskin, Krishna Sri Ipsit Mantri, Moshe Eliasof","submitted_at":"2025-02-09T21:05:11Z","abstract_excerpt":"Pre-trained Vision Transformers now serve as powerful tools for computer vision. Yet, efficiently adapting them for multiple tasks remains a challenge that arises from the need to modify the rich hidden representations encoded by the learned weight matrices, without inducing interference between tasks. Current parameter-efficient methods like LoRA, which apply low-rank updates, force tasks to compete within constrained subspaces, ultimately degrading performance. We introduce DiTASK a novel Diffeomorphic Multi-Task Fine-Tuning approach that maintains pre-trained representations by preserving w"},"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":"2502.06029","kind":"arxiv","version":3},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CV","submitted_at":"2025-02-09T21:05:11Z","cross_cats_sorted":[],"title_canon_sha256":"aa1b106b5129aef874d92870e3971b1e34a94984f6a968533afd975c22e757b0","abstract_canon_sha256":"b34bfb8ae63c8f083978a17b5158fe46837bb1f69a20eafd3cd85f63414c6cef"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T11:13:20.096192Z","signature_b64":"opelj1cSI4K1I0u0alpJ8eJ4diqf3eHujyS19C0dRQBZMe9jBVYRrk9jAANyiTTK1pb1nhQhwaQugXKETEBhBg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"3b10102d00938c1166b3cec42323aebb2c4dca28d92941bf8163f9cf4c8b7374","last_reissued_at":"2026-07-05T11:13:20.095732Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T11:13:20.095732Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"DiTASK: Multi-Task Fine-Tuning with Diffeomorphic Transformations","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.CV","authors_text":"Bruno Ribeiro, Carola-Bibiane Sch\\\"onlieb, Chaim Baskin, Krishna Sri Ipsit Mantri, Moshe Eliasof","submitted_at":"2025-02-09T21:05:11Z","abstract_excerpt":"Pre-trained Vision Transformers now serve as powerful tools for computer vision. Yet, efficiently adapting them for multiple tasks remains a challenge that arises from the need to modify the rich hidden representations encoded by the learned weight matrices, without inducing interference between tasks. Current parameter-efficient methods like LoRA, which apply low-rank updates, force tasks to compete within constrained subspaces, ultimately degrading performance. We introduce DiTASK a novel Diffeomorphic Multi-Task Fine-Tuning approach that maintains pre-trained representations by preserving w"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2502.06029","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/2502.06029/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":"2502.06029","created_at":"2026-07-05T11:13:20.095794+00:00"},{"alias_kind":"arxiv_version","alias_value":"2502.06029v3","created_at":"2026-07-05T11:13:20.095794+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2502.06029","created_at":"2026-07-05T11:13:20.095794+00:00"},{"alias_kind":"pith_short_12","alias_value":"HMIBALIASOGB","created_at":"2026-07-05T11:13:20.095794+00:00"},{"alias_kind":"pith_short_16","alias_value":"HMIBALIASOGBCZVT","created_at":"2026-07-05T11:13:20.095794+00:00"},{"alias_kind":"pith_short_8","alias_value":"HMIBALIA","created_at":"2026-07-05T11:13:20.095794+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/HMIBALIASOGBCZVTZ3CCGI5OXM","json":"https://pith.science/pith/HMIBALIASOGBCZVTZ3CCGI5OXM.json","graph_json":"https://pith.science/api/pith-number/HMIBALIASOGBCZVTZ3CCGI5OXM/graph.json","events_json":"https://pith.science/api/pith-number/HMIBALIASOGBCZVTZ3CCGI5OXM/events.json","paper":"https://pith.science/paper/HMIBALIA"},"agent_actions":{"view_html":"https://pith.science/pith/HMIBALIASOGBCZVTZ3CCGI5OXM","download_json":"https://pith.science/pith/HMIBALIASOGBCZVTZ3CCGI5OXM.json","view_paper":"https://pith.science/paper/HMIBALIA","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2502.06029&json=true","fetch_graph":"https://pith.science/api/pith-number/HMIBALIASOGBCZVTZ3CCGI5OXM/graph.json","fetch_events":"https://pith.science/api/pith-number/HMIBALIASOGBCZVTZ3CCGI5OXM/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/HMIBALIASOGBCZVTZ3CCGI5OXM/action/timestamp_anchor","attest_storage":"https://pith.science/pith/HMIBALIASOGBCZVTZ3CCGI5OXM/action/storage_attestation","attest_author":"https://pith.science/pith/HMIBALIASOGBCZVTZ3CCGI5OXM/action/author_attestation","sign_citation":"https://pith.science/pith/HMIBALIASOGBCZVTZ3CCGI5OXM/action/citation_signature","submit_replication":"https://pith.science/pith/HMIBALIASOGBCZVTZ3CCGI5OXM/action/replication_record"}},"created_at":"2026-07-05T11:13:20.095794+00:00","updated_at":"2026-07-05T11:13:20.095794+00:00"}