{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2021:YM372FHBVGCCJLJUINL77ZP72R","short_pith_number":"pith:YM372FHB","schema_version":"1.0","canonical_sha256":"c337fd14e1a98424ad344357ffe5ffd4472d5406acb20d81177d357418b2cd63","source":{"kind":"arxiv","id":"2110.08534","version":3},"attestation_state":"computed","paper":{"title":"Lifelong Pretraining: Continually Adapting Language Models to Emerging Corpora","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.CL","authors_text":"Andrew Arnold, Dejiao Zhang, Henghui Zhu, Shang-Wen Li, Wei Xiao, Xiang Ren, Xiaokai Wei, Xisen Jin","submitted_at":"2021-10-16T09:59:33Z","abstract_excerpt":"Pretrained language models (PTLMs) are typically learned over a large, static corpus and further fine-tuned for various downstream tasks. However, when deployed in the real world, a PTLM-based model must deal with data distributions that deviate from what the PTLM was initially trained on. In this paper, we study a lifelong language model pretraining challenge where a PTLM is continually updated so as to adapt to emerging data. Over a domain-incremental research paper stream and a chronologically-ordered tweet stream, we incrementally pretrain a PTLM with different continual learning algorithm"},"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":"2110.08534","kind":"arxiv","version":3},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2021-10-16T09:59:33Z","cross_cats_sorted":[],"title_canon_sha256":"7fda81c287e10c56f978b348ccbd50f9a6e1dbffef9a0d176e4418da4d508b74","abstract_canon_sha256":"7e099bfb43f5ef4224b61bb9513bd35ba1bb85e7b18eaa08736f654878c2bcf2"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T04:41:10.310392Z","signature_b64":"Ao2Ky7tEo5suKIirxUBClsSJwq0jxd5uez80OP0EMrA+D1JgauZwV1Dj7IT7yow4FPZQ6QomQkkEsPfaMuquBw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"c337fd14e1a98424ad344357ffe5ffd4472d5406acb20d81177d357418b2cd63","last_reissued_at":"2026-07-05T04:41:10.309908Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T04:41:10.309908Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Lifelong Pretraining: Continually Adapting Language Models to Emerging Corpora","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.CL","authors_text":"Andrew Arnold, Dejiao Zhang, Henghui Zhu, Shang-Wen Li, Wei Xiao, Xiang Ren, Xiaokai Wei, Xisen Jin","submitted_at":"2021-10-16T09:59:33Z","abstract_excerpt":"Pretrained language models (PTLMs) are typically learned over a large, static corpus and further fine-tuned for various downstream tasks. However, when deployed in the real world, a PTLM-based model must deal with data distributions that deviate from what the PTLM was initially trained on. In this paper, we study a lifelong language model pretraining challenge where a PTLM is continually updated so as to adapt to emerging data. Over a domain-incremental research paper stream and a chronologically-ordered tweet stream, we incrementally pretrain a PTLM with different continual learning algorithm"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2110.08534","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/2110.08534/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":"2110.08534","created_at":"2026-07-05T04:41:10.309964+00:00"},{"alias_kind":"arxiv_version","alias_value":"2110.08534v3","created_at":"2026-07-05T04:41:10.309964+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2110.08534","created_at":"2026-07-05T04:41:10.309964+00:00"},{"alias_kind":"pith_short_12","alias_value":"YM372FHBVGCC","created_at":"2026-07-05T04:41:10.309964+00:00"},{"alias_kind":"pith_short_16","alias_value":"YM372FHBVGCCJLJU","created_at":"2026-07-05T04:41:10.309964+00:00"},{"alias_kind":"pith_short_8","alias_value":"YM372FHB","created_at":"2026-07-05T04:41:10.309964+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":3,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2509.07177","citing_title":"Towards EnergyGPT: A Large Language Model Specialized for the Energy Sector","ref_index":19,"is_internal_anchor":false},{"citing_arxiv_id":"2401.05459","citing_title":"Personal LLM Agents: Insights and Survey about the Capability, Efficiency and Security","ref_index":239,"is_internal_anchor":false},{"citing_arxiv_id":"2305.14233","citing_title":"Enhancing Chat Language Models by Scaling High-quality Instructional Conversations","ref_index":185,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/YM372FHBVGCCJLJUINL77ZP72R","json":"https://pith.science/pith/YM372FHBVGCCJLJUINL77ZP72R.json","graph_json":"https://pith.science/api/pith-number/YM372FHBVGCCJLJUINL77ZP72R/graph.json","events_json":"https://pith.science/api/pith-number/YM372FHBVGCCJLJUINL77ZP72R/events.json","paper":"https://pith.science/paper/YM372FHB"},"agent_actions":{"view_html":"https://pith.science/pith/YM372FHBVGCCJLJUINL77ZP72R","download_json":"https://pith.science/pith/YM372FHBVGCCJLJUINL77ZP72R.json","view_paper":"https://pith.science/paper/YM372FHB","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2110.08534&json=true","fetch_graph":"https://pith.science/api/pith-number/YM372FHBVGCCJLJUINL77ZP72R/graph.json","fetch_events":"https://pith.science/api/pith-number/YM372FHBVGCCJLJUINL77ZP72R/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/YM372FHBVGCCJLJUINL77ZP72R/action/timestamp_anchor","attest_storage":"https://pith.science/pith/YM372FHBVGCCJLJUINL77ZP72R/action/storage_attestation","attest_author":"https://pith.science/pith/YM372FHBVGCCJLJUINL77ZP72R/action/author_attestation","sign_citation":"https://pith.science/pith/YM372FHBVGCCJLJUINL77ZP72R/action/citation_signature","submit_replication":"https://pith.science/pith/YM372FHBVGCCJLJUINL77ZP72R/action/replication_record"}},"created_at":"2026-07-05T04:41:10.309964+00:00","updated_at":"2026-07-05T04:41:10.309964+00:00"}