{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2024:KEBRURJ76WLIDSQBBKGHX4ZGD5","short_pith_number":"pith:KEBRURJ7","schema_version":"1.0","canonical_sha256":"51031a453ff59681ca010a8c7bf3261f5cc559be10f03399cd5879aa7ff9caaf","source":{"kind":"arxiv","id":"2405.18756","version":1},"attestation_state":"computed","paper":{"title":"Provable Contrastive Continual Learning","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.AI","cs.CV","stat.AP","stat.ML"],"primary_cat":"cs.LG","authors_text":"Chuanlong Xie, Kaipeng Zheng, Weiran Huang, Yichen Wen, Zhiquan Tan","submitted_at":"2024-05-29T04:48:11Z","abstract_excerpt":"Continual learning requires learning incremental tasks with dynamic data distributions. So far, it has been observed that employing a combination of contrastive loss and distillation loss for training in continual learning yields strong performance. To the best of our knowledge, however, this contrastive continual learning framework lacks convincing theoretical explanations. In this work, we fill this gap by establishing theoretical performance guarantees, which reveal how the performance of the model is bounded by training losses of previous tasks in the contrastive continual learning framewo"},"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":"2405.18756","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2024-05-29T04:48:11Z","cross_cats_sorted":["cs.AI","cs.CV","stat.AP","stat.ML"],"title_canon_sha256":"04d3f29d0781e4dfe8bdb8a238f857708fe78b6048053bdb422951d33d9c78f9","abstract_canon_sha256":"850a611ff56350a1ce393a333055af3819fbf7608599730e3b6486b141f87280"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T08:24:29.873241Z","signature_b64":"Vy4ZMbboo7w5UXszqxl4rvtXZyeuUqXuyAxkWkvQ6AjR/nOqDkB5NoFR4hKBkfRhsUrSZ35fvJIJQtnqdO33Dg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"51031a453ff59681ca010a8c7bf3261f5cc559be10f03399cd5879aa7ff9caaf","last_reissued_at":"2026-07-05T08:24:29.872769Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T08:24:29.872769Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Provable Contrastive Continual Learning","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.AI","cs.CV","stat.AP","stat.ML"],"primary_cat":"cs.LG","authors_text":"Chuanlong Xie, Kaipeng Zheng, Weiran Huang, Yichen Wen, Zhiquan Tan","submitted_at":"2024-05-29T04:48:11Z","abstract_excerpt":"Continual learning requires learning incremental tasks with dynamic data distributions. So far, it has been observed that employing a combination of contrastive loss and distillation loss for training in continual learning yields strong performance. To the best of our knowledge, however, this contrastive continual learning framework lacks convincing theoretical explanations. In this work, we fill this gap by establishing theoretical performance guarantees, which reveal how the performance of the model is bounded by training losses of previous tasks in the contrastive continual learning framewo"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2405.18756","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/2405.18756/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":"2405.18756","created_at":"2026-07-05T08:24:29.872826+00:00"},{"alias_kind":"arxiv_version","alias_value":"2405.18756v1","created_at":"2026-07-05T08:24:29.872826+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2405.18756","created_at":"2026-07-05T08:24:29.872826+00:00"},{"alias_kind":"pith_short_12","alias_value":"KEBRURJ76WLI","created_at":"2026-07-05T08:24:29.872826+00:00"},{"alias_kind":"pith_short_16","alias_value":"KEBRURJ76WLIDSQB","created_at":"2026-07-05T08:24:29.872826+00:00"},{"alias_kind":"pith_short_8","alias_value":"KEBRURJ7","created_at":"2026-07-05T08:24:29.872826+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":1,"internal_anchor_count":1,"sample":[{"citing_arxiv_id":"2505.20970","citing_title":"Measuring Representational Shifts in Continual Learning: A Linear Transformation Perspective","ref_index":43,"is_internal_anchor":true}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/KEBRURJ76WLIDSQBBKGHX4ZGD5","json":"https://pith.science/pith/KEBRURJ76WLIDSQBBKGHX4ZGD5.json","graph_json":"https://pith.science/api/pith-number/KEBRURJ76WLIDSQBBKGHX4ZGD5/graph.json","events_json":"https://pith.science/api/pith-number/KEBRURJ76WLIDSQBBKGHX4ZGD5/events.json","paper":"https://pith.science/paper/KEBRURJ7"},"agent_actions":{"view_html":"https://pith.science/pith/KEBRURJ76WLIDSQBBKGHX4ZGD5","download_json":"https://pith.science/pith/KEBRURJ76WLIDSQBBKGHX4ZGD5.json","view_paper":"https://pith.science/paper/KEBRURJ7","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2405.18756&json=true","fetch_graph":"https://pith.science/api/pith-number/KEBRURJ76WLIDSQBBKGHX4ZGD5/graph.json","fetch_events":"https://pith.science/api/pith-number/KEBRURJ76WLIDSQBBKGHX4ZGD5/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/KEBRURJ76WLIDSQBBKGHX4ZGD5/action/timestamp_anchor","attest_storage":"https://pith.science/pith/KEBRURJ76WLIDSQBBKGHX4ZGD5/action/storage_attestation","attest_author":"https://pith.science/pith/KEBRURJ76WLIDSQBBKGHX4ZGD5/action/author_attestation","sign_citation":"https://pith.science/pith/KEBRURJ76WLIDSQBBKGHX4ZGD5/action/citation_signature","submit_replication":"https://pith.science/pith/KEBRURJ76WLIDSQBBKGHX4ZGD5/action/replication_record"}},"created_at":"2026-07-05T08:24:29.872826+00:00","updated_at":"2026-07-05T08:24:29.872826+00:00"}