{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2023:PVJ4KDXMYMCSBE3JGNRACESC22","short_pith_number":"pith:PVJ4KDXM","schema_version":"1.0","canonical_sha256":"7d53c50eecc3052093693362011242d6966ce202ff6c824331ba50760d27f9d4","source":{"kind":"arxiv","id":"2304.14381","version":3},"attestation_state":"computed","paper":{"title":"$\\pi$-Tuning: Transferring Multimodal Foundation Models with Optimal Multi-task Interpolation","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.CV","authors_text":"Chengyue Wu, Ping Luo, Ruisong Zhou, Teng Wang, Ying Shan, Yixiao Ge, Zeyu Lu","submitted_at":"2023-04-27T17:49:54Z","abstract_excerpt":"Foundation models have achieved great advances in multi-task learning with a unified interface of unimodal and multimodal tasks. However, the potential of such multi-task learners has not been exploited during transfer learning. In this work, we present a universal parameter-efficient transfer learning method, termed Predict-Interpolate Tuning ($\\pi$-Tuning), for vision, language, and vision-language tasks. It aggregates the parameters of lightweight task-specific experts learned from similar tasks to aid the target downstream task. The task similarities are predicted in a unified modality-ind"},"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":"2304.14381","kind":"arxiv","version":3},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CV","submitted_at":"2023-04-27T17:49:54Z","cross_cats_sorted":[],"title_canon_sha256":"3ffb50cc49d30971305937280a44385a15add54da0593bb8a148edbdb45ad97e","abstract_canon_sha256":"ce3cb29c806c02825e5dc10b991513cc5e9b4bb5ecbc26b63a1b6ccca5c1b133"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T06:11:04.589184Z","signature_b64":"1lN0a5LKgqungtTDg82UcF+c6zTf4iZU84zE+oqtPuyOQwVBJnS2/VzdKInNWu9Taot6Oukv+VYXx9vJ2GeBDw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"7d53c50eecc3052093693362011242d6966ce202ff6c824331ba50760d27f9d4","last_reissued_at":"2026-07-05T06:11:04.588709Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T06:11:04.588709Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"$\\pi$-Tuning: Transferring Multimodal Foundation Models with Optimal Multi-task Interpolation","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.CV","authors_text":"Chengyue Wu, Ping Luo, Ruisong Zhou, Teng Wang, Ying Shan, Yixiao Ge, Zeyu Lu","submitted_at":"2023-04-27T17:49:54Z","abstract_excerpt":"Foundation models have achieved great advances in multi-task learning with a unified interface of unimodal and multimodal tasks. However, the potential of such multi-task learners has not been exploited during transfer learning. In this work, we present a universal parameter-efficient transfer learning method, termed Predict-Interpolate Tuning ($\\pi$-Tuning), for vision, language, and vision-language tasks. It aggregates the parameters of lightweight task-specific experts learned from similar tasks to aid the target downstream task. The task similarities are predicted in a unified modality-ind"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2304.14381","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/2304.14381/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":"2304.14381","created_at":"2026-07-05T06:11:04.588771+00:00"},{"alias_kind":"arxiv_version","alias_value":"2304.14381v3","created_at":"2026-07-05T06:11:04.588771+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2304.14381","created_at":"2026-07-05T06:11:04.588771+00:00"},{"alias_kind":"pith_short_12","alias_value":"PVJ4KDXMYMCS","created_at":"2026-07-05T06:11:04.588771+00:00"},{"alias_kind":"pith_short_16","alias_value":"PVJ4KDXMYMCSBE3J","created_at":"2026-07-05T06:11:04.588771+00:00"},{"alias_kind":"pith_short_8","alias_value":"PVJ4KDXM","created_at":"2026-07-05T06:11:04.588771+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/PVJ4KDXMYMCSBE3JGNRACESC22","json":"https://pith.science/pith/PVJ4KDXMYMCSBE3JGNRACESC22.json","graph_json":"https://pith.science/api/pith-number/PVJ4KDXMYMCSBE3JGNRACESC22/graph.json","events_json":"https://pith.science/api/pith-number/PVJ4KDXMYMCSBE3JGNRACESC22/events.json","paper":"https://pith.science/paper/PVJ4KDXM"},"agent_actions":{"view_html":"https://pith.science/pith/PVJ4KDXMYMCSBE3JGNRACESC22","download_json":"https://pith.science/pith/PVJ4KDXMYMCSBE3JGNRACESC22.json","view_paper":"https://pith.science/paper/PVJ4KDXM","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2304.14381&json=true","fetch_graph":"https://pith.science/api/pith-number/PVJ4KDXMYMCSBE3JGNRACESC22/graph.json","fetch_events":"https://pith.science/api/pith-number/PVJ4KDXMYMCSBE3JGNRACESC22/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/PVJ4KDXMYMCSBE3JGNRACESC22/action/timestamp_anchor","attest_storage":"https://pith.science/pith/PVJ4KDXMYMCSBE3JGNRACESC22/action/storage_attestation","attest_author":"https://pith.science/pith/PVJ4KDXMYMCSBE3JGNRACESC22/action/author_attestation","sign_citation":"https://pith.science/pith/PVJ4KDXMYMCSBE3JGNRACESC22/action/citation_signature","submit_replication":"https://pith.science/pith/PVJ4KDXMYMCSBE3JGNRACESC22/action/replication_record"}},"created_at":"2026-07-05T06:11:04.588771+00:00","updated_at":"2026-07-05T06:11:04.588771+00:00"}