{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2024:NQAX22Z3R6KRIQN6WVGZQJWSFL","short_pith_number":"pith:NQAX22Z3","schema_version":"1.0","canonical_sha256":"6c017d6b3b8f951441beb54d9826d22ae707b2c721b3643b76c1edb1760312ae","source":{"kind":"arxiv","id":"2401.15275","version":1},"attestation_state":"computed","paper":{"title":"Dynamic Transformer Architecture for Continual Learning of Multimodal Tasks","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.CV","authors_text":"Mohammad Rostami, Yuliang Cai","submitted_at":"2024-01-27T03:03:30Z","abstract_excerpt":"Transformer neural networks are increasingly replacing prior architectures in a wide range of applications in different data modalities. The increasing size and computational demands of fine-tuning large pre-trained transformer neural networks pose significant challenges for the widespread adoption of these models for applications that demand on-edge computing. To tackle this challenge, continual learning (CL) emerges as a solution by facilitating the transfer of knowledge across tasks that arrive sequentially for an autonomously learning agent. However, current CL methods mainly focus on lear"},"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":"2401.15275","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CV","submitted_at":"2024-01-27T03:03:30Z","cross_cats_sorted":[],"title_canon_sha256":"63120f60ac141278f8c07671717ddf236ac431cd9d3f9343f7bc4c81bbe13277","abstract_canon_sha256":"8629a25a329da2fb0b0c1773ae17abc0b60b95d8d110a7cffb3466ce03da5b71"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T07:38:13.787673Z","signature_b64":"h56tkPeGHEoDiWPfdrKMwCH/zsgAS0M/rEAxt+TMncU6jCvV4cflpu2kw4dyRmaQGwOvlT4hjQ4w4ehtdelZBw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"6c017d6b3b8f951441beb54d9826d22ae707b2c721b3643b76c1edb1760312ae","last_reissued_at":"2026-07-05T07:38:13.787241Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T07:38:13.787241Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Dynamic Transformer Architecture for Continual Learning of Multimodal Tasks","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.CV","authors_text":"Mohammad Rostami, Yuliang Cai","submitted_at":"2024-01-27T03:03:30Z","abstract_excerpt":"Transformer neural networks are increasingly replacing prior architectures in a wide range of applications in different data modalities. The increasing size and computational demands of fine-tuning large pre-trained transformer neural networks pose significant challenges for the widespread adoption of these models for applications that demand on-edge computing. To tackle this challenge, continual learning (CL) emerges as a solution by facilitating the transfer of knowledge across tasks that arrive sequentially for an autonomously learning agent. However, current CL methods mainly focus on lear"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2401.15275","kind":"arxiv","version":1},"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/2401.15275/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":"2401.15275","created_at":"2026-07-05T07:38:13.787299+00:00"},{"alias_kind":"arxiv_version","alias_value":"2401.15275v1","created_at":"2026-07-05T07:38:13.787299+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2401.15275","created_at":"2026-07-05T07:38:13.787299+00:00"},{"alias_kind":"pith_short_12","alias_value":"NQAX22Z3R6KR","created_at":"2026-07-05T07:38:13.787299+00:00"},{"alias_kind":"pith_short_16","alias_value":"NQAX22Z3R6KRIQN6","created_at":"2026-07-05T07:38:13.787299+00:00"},{"alias_kind":"pith_short_8","alias_value":"NQAX22Z3","created_at":"2026-07-05T07:38:13.787299+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":1,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2605.05495","citing_title":"Shortcut Solutions Learned by Transformers Impair Continual Compositional Reasoning","ref_index":4,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/NQAX22Z3R6KRIQN6WVGZQJWSFL","json":"https://pith.science/pith/NQAX22Z3R6KRIQN6WVGZQJWSFL.json","graph_json":"https://pith.science/api/pith-number/NQAX22Z3R6KRIQN6WVGZQJWSFL/graph.json","events_json":"https://pith.science/api/pith-number/NQAX22Z3R6KRIQN6WVGZQJWSFL/events.json","paper":"https://pith.science/paper/NQAX22Z3"},"agent_actions":{"view_html":"https://pith.science/pith/NQAX22Z3R6KRIQN6WVGZQJWSFL","download_json":"https://pith.science/pith/NQAX22Z3R6KRIQN6WVGZQJWSFL.json","view_paper":"https://pith.science/paper/NQAX22Z3","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2401.15275&json=true","fetch_graph":"https://pith.science/api/pith-number/NQAX22Z3R6KRIQN6WVGZQJWSFL/graph.json","fetch_events":"https://pith.science/api/pith-number/NQAX22Z3R6KRIQN6WVGZQJWSFL/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/NQAX22Z3R6KRIQN6WVGZQJWSFL/action/timestamp_anchor","attest_storage":"https://pith.science/pith/NQAX22Z3R6KRIQN6WVGZQJWSFL/action/storage_attestation","attest_author":"https://pith.science/pith/NQAX22Z3R6KRIQN6WVGZQJWSFL/action/author_attestation","sign_citation":"https://pith.science/pith/NQAX22Z3R6KRIQN6WVGZQJWSFL/action/citation_signature","submit_replication":"https://pith.science/pith/NQAX22Z3R6KRIQN6WVGZQJWSFL/action/replication_record"}},"created_at":"2026-07-05T07:38:13.787299+00:00","updated_at":"2026-07-05T07:38:13.787299+00:00"}