{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2025:N2PG4VLZIHOEDLYTL6I5BFLTEF","short_pith_number":"pith:N2PG4VLZ","schema_version":"1.0","canonical_sha256":"6e9e6e557941dc41af135f91d095732145e28fe4501d09bc55c201e94ab9641c","source":{"kind":"arxiv","id":"2508.07738","version":1},"attestation_state":"computed","paper":{"title":"Separation and Collaboration: Two-Level Routing Grouped Mixture-of-Experts for Multi-Domain Continual Learning","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.LG","authors_text":"Chunping Qiu, Dianxi Shi, Jialu Zhou, Leqian Li, Mengzhu Wang, Mingyue Yang, Shaowu Yang, Xinyu Wei","submitted_at":"2025-08-11T08:18:22Z","abstract_excerpt":"Multi-Domain Continual Learning (MDCL) acquires knowledge from sequential tasks with shifting class sets and distribution. Despite the Parameter-Efficient Fine-Tuning (PEFT) methods can adapt for this dual heterogeneity, they still suffer from catastrophic forgetting and forward forgetting. To address these challenges, we propose a Two-Level Routing Grouped Mixture-of-Experts (TRGE) method. Firstly, TRGE dynamically expands the pre-trained CLIP model, assigning specific expert group for each task to mitigate catastrophic forgetting. With the number of experts continually grows in this process,"},"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":"2508.07738","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2025-08-11T08:18:22Z","cross_cats_sorted":[],"title_canon_sha256":"1b542b804ee440a0860fffcaaf789d232d8bca5d26a91eba821c4e82c6dfa395","abstract_canon_sha256":"b84ca7de21aa6d8761ec6040b8eaa09bd07ac436cc97ede781758f845d56eba4"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T11:51:51.322168Z","signature_b64":"+GAJRtVijgXcHN9TAOPSAgRv0XzXRrU/omQ2VpRgOzgWwWLBUQQ3zlUgiHbdP4r1/a7jlm3JWOyY7+Mox7T9Aw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"6e9e6e557941dc41af135f91d095732145e28fe4501d09bc55c201e94ab9641c","last_reissued_at":"2026-07-05T11:51:51.321727Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T11:51:51.321727Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Separation and Collaboration: Two-Level Routing Grouped Mixture-of-Experts for Multi-Domain Continual Learning","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.LG","authors_text":"Chunping Qiu, Dianxi Shi, Jialu Zhou, Leqian Li, Mengzhu Wang, Mingyue Yang, Shaowu Yang, Xinyu Wei","submitted_at":"2025-08-11T08:18:22Z","abstract_excerpt":"Multi-Domain Continual Learning (MDCL) acquires knowledge from sequential tasks with shifting class sets and distribution. Despite the Parameter-Efficient Fine-Tuning (PEFT) methods can adapt for this dual heterogeneity, they still suffer from catastrophic forgetting and forward forgetting. To address these challenges, we propose a Two-Level Routing Grouped Mixture-of-Experts (TRGE) method. Firstly, TRGE dynamically expands the pre-trained CLIP model, assigning specific expert group for each task to mitigate catastrophic forgetting. With the number of experts continually grows in this process,"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2508.07738","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/2508.07738/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":"2508.07738","created_at":"2026-07-05T11:51:51.321783+00:00"},{"alias_kind":"arxiv_version","alias_value":"2508.07738v1","created_at":"2026-07-05T11:51:51.321783+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2508.07738","created_at":"2026-07-05T11:51:51.321783+00:00"},{"alias_kind":"pith_short_12","alias_value":"N2PG4VLZIHOE","created_at":"2026-07-05T11:51:51.321783+00:00"},{"alias_kind":"pith_short_16","alias_value":"N2PG4VLZIHOEDLYT","created_at":"2026-07-05T11:51:51.321783+00:00"},{"alias_kind":"pith_short_8","alias_value":"N2PG4VLZ","created_at":"2026-07-05T11:51:51.321783+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":1,"internal_anchor_count":1,"sample":[{"citing_arxiv_id":"2508.07723","citing_title":"Enhancing Small-Scale Dataset Expansion with Triplet-Connection-based Sample Re-Weighting","ref_index":2021,"is_internal_anchor":true}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/N2PG4VLZIHOEDLYTL6I5BFLTEF","json":"https://pith.science/pith/N2PG4VLZIHOEDLYTL6I5BFLTEF.json","graph_json":"https://pith.science/api/pith-number/N2PG4VLZIHOEDLYTL6I5BFLTEF/graph.json","events_json":"https://pith.science/api/pith-number/N2PG4VLZIHOEDLYTL6I5BFLTEF/events.json","paper":"https://pith.science/paper/N2PG4VLZ"},"agent_actions":{"view_html":"https://pith.science/pith/N2PG4VLZIHOEDLYTL6I5BFLTEF","download_json":"https://pith.science/pith/N2PG4VLZIHOEDLYTL6I5BFLTEF.json","view_paper":"https://pith.science/paper/N2PG4VLZ","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2508.07738&json=true","fetch_graph":"https://pith.science/api/pith-number/N2PG4VLZIHOEDLYTL6I5BFLTEF/graph.json","fetch_events":"https://pith.science/api/pith-number/N2PG4VLZIHOEDLYTL6I5BFLTEF/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/N2PG4VLZIHOEDLYTL6I5BFLTEF/action/timestamp_anchor","attest_storage":"https://pith.science/pith/N2PG4VLZIHOEDLYTL6I5BFLTEF/action/storage_attestation","attest_author":"https://pith.science/pith/N2PG4VLZIHOEDLYTL6I5BFLTEF/action/author_attestation","sign_citation":"https://pith.science/pith/N2PG4VLZIHOEDLYTL6I5BFLTEF/action/citation_signature","submit_replication":"https://pith.science/pith/N2PG4VLZIHOEDLYTL6I5BFLTEF/action/replication_record"}},"created_at":"2026-07-05T11:51:51.321783+00:00","updated_at":"2026-07-05T11:51:51.321783+00:00"}