{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2021:IPM4TSFKXHT6DQ56G5HA2SKB7K","short_pith_number":"pith:IPM4TSFK","schema_version":"1.0","canonical_sha256":"43d9c9c8aab9e7e1c3be374e0d4941faa15c4ea135041ae53460759354a70de5","source":{"kind":"arxiv","id":"2101.08751","version":1},"attestation_state":"computed","paper":{"title":"Rethink Training of BERT Rerankers in Multi-Stage Retrieval Pipeline","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.IR","authors_text":"Jamie Callan, Luyu Gao, Zhuyun Dai","submitted_at":"2021-01-21T18:04:58Z","abstract_excerpt":"Pre-trained deep language models~(LM) have advanced the state-of-the-art of text retrieval. Rerankers fine-tuned from deep LM estimates candidate relevance based on rich contextualized matching signals. Meanwhile, deep LMs can also be leveraged to improve search index, building retrievers with better recall. One would expect a straightforward combination of both in a pipeline to have additive performance gain. In this paper, we discover otherwise and that popular reranker cannot fully exploit the improved retrieval result. We, therefore, propose a Localized Contrastive Estimation (LCE) for tra"},"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":"2101.08751","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.IR","submitted_at":"2021-01-21T18:04:58Z","cross_cats_sorted":[],"title_canon_sha256":"2c973b580a23f4bc2f8db90f49bc070f3a9af3ebb31c245ff71ffa114b2e36b5","abstract_canon_sha256":"f8e3761b4a5f3587d14d6564495287c5e41139038e79fd964198e7e10eebe0e0"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T02:08:38.277400Z","signature_b64":"MDWd/dvjlIOIxlEvWca9XUtByMs2VchkRzYghVfR2jgR0rOlxFVu3W3MMfsNNpFEWhnFxVcnLXjeBFuDRjQnBw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"43d9c9c8aab9e7e1c3be374e0d4941faa15c4ea135041ae53460759354a70de5","last_reissued_at":"2026-07-05T02:08:38.277039Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T02:08:38.277039Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Rethink Training of BERT Rerankers in Multi-Stage Retrieval Pipeline","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.IR","authors_text":"Jamie Callan, Luyu Gao, Zhuyun Dai","submitted_at":"2021-01-21T18:04:58Z","abstract_excerpt":"Pre-trained deep language models~(LM) have advanced the state-of-the-art of text retrieval. Rerankers fine-tuned from deep LM estimates candidate relevance based on rich contextualized matching signals. Meanwhile, deep LMs can also be leveraged to improve search index, building retrievers with better recall. One would expect a straightforward combination of both in a pipeline to have additive performance gain. In this paper, we discover otherwise and that popular reranker cannot fully exploit the improved retrieval result. We, therefore, propose a Localized Contrastive Estimation (LCE) for tra"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2101.08751","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/2101.08751/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":"2101.08751","created_at":"2026-07-05T02:08:38.277104+00:00"},{"alias_kind":"arxiv_version","alias_value":"2101.08751v1","created_at":"2026-07-05T02:08:38.277104+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2101.08751","created_at":"2026-07-05T02:08:38.277104+00:00"},{"alias_kind":"pith_short_12","alias_value":"IPM4TSFKXHT6","created_at":"2026-07-05T02:08:38.277104+00:00"},{"alias_kind":"pith_short_16","alias_value":"IPM4TSFKXHT6DQ56","created_at":"2026-07-05T02:08:38.277104+00:00"},{"alias_kind":"pith_short_8","alias_value":"IPM4TSFK","created_at":"2026-07-05T02:08:38.277104+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":1,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2604.26483","citing_title":"Efficient Listwise Reranking with Compressed Document Representations","ref_index":12,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/IPM4TSFKXHT6DQ56G5HA2SKB7K","json":"https://pith.science/pith/IPM4TSFKXHT6DQ56G5HA2SKB7K.json","graph_json":"https://pith.science/api/pith-number/IPM4TSFKXHT6DQ56G5HA2SKB7K/graph.json","events_json":"https://pith.science/api/pith-number/IPM4TSFKXHT6DQ56G5HA2SKB7K/events.json","paper":"https://pith.science/paper/IPM4TSFK"},"agent_actions":{"view_html":"https://pith.science/pith/IPM4TSFKXHT6DQ56G5HA2SKB7K","download_json":"https://pith.science/pith/IPM4TSFKXHT6DQ56G5HA2SKB7K.json","view_paper":"https://pith.science/paper/IPM4TSFK","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2101.08751&json=true","fetch_graph":"https://pith.science/api/pith-number/IPM4TSFKXHT6DQ56G5HA2SKB7K/graph.json","fetch_events":"https://pith.science/api/pith-number/IPM4TSFKXHT6DQ56G5HA2SKB7K/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/IPM4TSFKXHT6DQ56G5HA2SKB7K/action/timestamp_anchor","attest_storage":"https://pith.science/pith/IPM4TSFKXHT6DQ56G5HA2SKB7K/action/storage_attestation","attest_author":"https://pith.science/pith/IPM4TSFKXHT6DQ56G5HA2SKB7K/action/author_attestation","sign_citation":"https://pith.science/pith/IPM4TSFKXHT6DQ56G5HA2SKB7K/action/citation_signature","submit_replication":"https://pith.science/pith/IPM4TSFKXHT6DQ56G5HA2SKB7K/action/replication_record"}},"created_at":"2026-07-05T02:08:38.277104+00:00","updated_at":"2026-07-05T02:08:38.277104+00:00"}