{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2021:4KTTRKTKBHPEUEPTHPKNFEAJPZ","short_pith_number":"pith:4KTTRKTK","schema_version":"1.0","canonical_sha256":"e2a738aa6a09de4a11f33bd4d290097e515f15a9a3d7c1ac79106ea2d1d3659c","source":{"kind":"arxiv","id":"2109.04810","version":1},"attestation_state":"computed","paper":{"title":"Mixture-of-Partitions: Infusing Large Biomedical Knowledge Graphs into BERT","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.CL","authors_text":"Ehsan Shareghi, Fangyu Liu, Nigel Collier, Thomas Hikaru Clark, Zaiqiao Meng","submitted_at":"2021-09-10T11:54:25Z","abstract_excerpt":"Infusing factual knowledge into pre-trained models is fundamental for many knowledge-intensive tasks. In this paper, we proposed Mixture-of-Partitions (MoP), an infusion approach that can handle a very large knowledge graph (KG) by partitioning it into smaller sub-graphs and infusing their specific knowledge into various BERT models using lightweight adapters. To leverage the overall factual knowledge for a target task, these sub-graph adapters are further fine-tuned along with the underlying BERT through a mixture layer. We evaluate our MoP with three biomedical BERTs (SciBERT, BioBERT, Pubme"},"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":"2109.04810","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2021-09-10T11:54:25Z","cross_cats_sorted":[],"title_canon_sha256":"b6eacbca314665cd1ea9d29ae81ba8cd560e8a7b82c266adaefae81973eb49f4","abstract_canon_sha256":"d0457bc16a55e069d81448acf82dde21ff7ee885e4cb8a644a11ca2a2e6cba5f"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T03:13:10.392255Z","signature_b64":"da+O3HiagRWvMJBVv4fCI9GnZR1ihu+4FJhCrfcB+N2MlwIQCTCLVE9iAyOIlNHXXAnpL+sILiM+1MAK8sfKBw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"e2a738aa6a09de4a11f33bd4d290097e515f15a9a3d7c1ac79106ea2d1d3659c","last_reissued_at":"2026-07-05T03:13:10.391881Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T03:13:10.391881Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Mixture-of-Partitions: Infusing Large Biomedical Knowledge Graphs into BERT","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.CL","authors_text":"Ehsan Shareghi, Fangyu Liu, Nigel Collier, Thomas Hikaru Clark, Zaiqiao Meng","submitted_at":"2021-09-10T11:54:25Z","abstract_excerpt":"Infusing factual knowledge into pre-trained models is fundamental for many knowledge-intensive tasks. In this paper, we proposed Mixture-of-Partitions (MoP), an infusion approach that can handle a very large knowledge graph (KG) by partitioning it into smaller sub-graphs and infusing their specific knowledge into various BERT models using lightweight adapters. To leverage the overall factual knowledge for a target task, these sub-graph adapters are further fine-tuned along with the underlying BERT through a mixture layer. We evaluate our MoP with three biomedical BERTs (SciBERT, BioBERT, Pubme"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2109.04810","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/2109.04810/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":"2109.04810","created_at":"2026-07-05T03:13:10.391947+00:00"},{"alias_kind":"arxiv_version","alias_value":"2109.04810v1","created_at":"2026-07-05T03:13:10.391947+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2109.04810","created_at":"2026-07-05T03:13:10.391947+00:00"},{"alias_kind":"pith_short_12","alias_value":"4KTTRKTKBHPE","created_at":"2026-07-05T03:13:10.391947+00:00"},{"alias_kind":"pith_short_16","alias_value":"4KTTRKTKBHPEUEPT","created_at":"2026-07-05T03:13:10.391947+00:00"},{"alias_kind":"pith_short_8","alias_value":"4KTTRKTK","created_at":"2026-07-05T03:13:10.391947+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":1,"internal_anchor_count":1,"sample":[{"citing_arxiv_id":"2411.16403","citing_title":"Adapter-based Approaches to Knowledge-enhanced Language Models -- A Survey","ref_index":49,"is_internal_anchor":true}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/4KTTRKTKBHPEUEPTHPKNFEAJPZ","json":"https://pith.science/pith/4KTTRKTKBHPEUEPTHPKNFEAJPZ.json","graph_json":"https://pith.science/api/pith-number/4KTTRKTKBHPEUEPTHPKNFEAJPZ/graph.json","events_json":"https://pith.science/api/pith-number/4KTTRKTKBHPEUEPTHPKNFEAJPZ/events.json","paper":"https://pith.science/paper/4KTTRKTK"},"agent_actions":{"view_html":"https://pith.science/pith/4KTTRKTKBHPEUEPTHPKNFEAJPZ","download_json":"https://pith.science/pith/4KTTRKTKBHPEUEPTHPKNFEAJPZ.json","view_paper":"https://pith.science/paper/4KTTRKTK","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2109.04810&json=true","fetch_graph":"https://pith.science/api/pith-number/4KTTRKTKBHPEUEPTHPKNFEAJPZ/graph.json","fetch_events":"https://pith.science/api/pith-number/4KTTRKTKBHPEUEPTHPKNFEAJPZ/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/4KTTRKTKBHPEUEPTHPKNFEAJPZ/action/timestamp_anchor","attest_storage":"https://pith.science/pith/4KTTRKTKBHPEUEPTHPKNFEAJPZ/action/storage_attestation","attest_author":"https://pith.science/pith/4KTTRKTKBHPEUEPTHPKNFEAJPZ/action/author_attestation","sign_citation":"https://pith.science/pith/4KTTRKTKBHPEUEPTHPKNFEAJPZ/action/citation_signature","submit_replication":"https://pith.science/pith/4KTTRKTKBHPEUEPTHPKNFEAJPZ/action/replication_record"}},"created_at":"2026-07-05T03:13:10.391947+00:00","updated_at":"2026-07-05T03:13:10.391947+00:00"}