{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2022:HPUJUJ5LJ7GPWUDTTZ4YLXWXC3","short_pith_number":"pith:HPUJUJ5L","schema_version":"1.0","canonical_sha256":"3be89a27ab4fccfb50739e7985ded716e5744f586b1941e60e05230fb29aad26","source":{"kind":"arxiv","id":"2207.02578","version":2},"attestation_state":"computed","paper":{"title":"SimLM: Pre-training with Representation Bottleneck for Dense Passage Retrieval","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.IR","authors_text":"Binxing Jiao, Daxin Jiang, Furu Wei, Liang Wang, Linjun Yang, Nan Yang, Rangan Majumder, Xiaolong Huang","submitted_at":"2022-07-06T10:51:33Z","abstract_excerpt":"In this paper, we propose SimLM (Similarity matching with Language Model pre-training), a simple yet effective pre-training method for dense passage retrieval. It employs a simple bottleneck architecture that learns to compress the passage information into a dense vector through self-supervised pre-training. We use a replaced language modeling objective, which is inspired by ELECTRA, to improve the sample efficiency and reduce the mismatch of the input distribution between pre-training and fine-tuning. SimLM only requires access to unlabeled corpus, and is more broadly applicable when there ar"},"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":"2207.02578","kind":"arxiv","version":2},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.IR","submitted_at":"2022-07-06T10:51:33Z","cross_cats_sorted":[],"title_canon_sha256":"7cb08ff0427ea6fd139d7fa6342ef6ad90a83d5865a63a76d73b3d56701d85f0","abstract_canon_sha256":"bbfb41489fa2e6a2e581f2be1407e0c92409410bad8b07fbe88861e91f4ed45d"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T06:09:31.389760Z","signature_b64":"TMEkJm7BfKoqRV8kHuFdDO6RjpdzwJC1cmytM5JIPrUo+HqApD2Utul7yP3GXSDTxhNrlH2YWk+zH5/smGEmCA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"3be89a27ab4fccfb50739e7985ded716e5744f586b1941e60e05230fb29aad26","last_reissued_at":"2026-07-05T06:09:31.389340Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T06:09:31.389340Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"SimLM: Pre-training with Representation Bottleneck for Dense Passage Retrieval","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.IR","authors_text":"Binxing Jiao, Daxin Jiang, Furu Wei, Liang Wang, Linjun Yang, Nan Yang, Rangan Majumder, Xiaolong Huang","submitted_at":"2022-07-06T10:51:33Z","abstract_excerpt":"In this paper, we propose SimLM (Similarity matching with Language Model pre-training), a simple yet effective pre-training method for dense passage retrieval. It employs a simple bottleneck architecture that learns to compress the passage information into a dense vector through self-supervised pre-training. We use a replaced language modeling objective, which is inspired by ELECTRA, to improve the sample efficiency and reduce the mismatch of the input distribution between pre-training and fine-tuning. SimLM only requires access to unlabeled corpus, and is more broadly applicable when there ar"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2207.02578","kind":"arxiv","version":2},"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/2207.02578/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":"2207.02578","created_at":"2026-07-05T06:09:31.389396+00:00"},{"alias_kind":"arxiv_version","alias_value":"2207.02578v2","created_at":"2026-07-05T06:09:31.389396+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2207.02578","created_at":"2026-07-05T06:09:31.389396+00:00"},{"alias_kind":"pith_short_12","alias_value":"HPUJUJ5LJ7GP","created_at":"2026-07-05T06:09:31.389396+00:00"},{"alias_kind":"pith_short_16","alias_value":"HPUJUJ5LJ7GPWUDT","created_at":"2026-07-05T06:09:31.389396+00:00"},{"alias_kind":"pith_short_8","alias_value":"HPUJUJ5L","created_at":"2026-07-05T06:09:31.389396+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":2,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2309.07597","citing_title":"C-Pack: Packed Resources For General Chinese Embeddings","ref_index":57,"is_internal_anchor":false},{"citing_arxiv_id":"2212.03533","citing_title":"Text Embeddings by Weakly-Supervised Contrastive Pre-training","ref_index":61,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/HPUJUJ5LJ7GPWUDTTZ4YLXWXC3","json":"https://pith.science/pith/HPUJUJ5LJ7GPWUDTTZ4YLXWXC3.json","graph_json":"https://pith.science/api/pith-number/HPUJUJ5LJ7GPWUDTTZ4YLXWXC3/graph.json","events_json":"https://pith.science/api/pith-number/HPUJUJ5LJ7GPWUDTTZ4YLXWXC3/events.json","paper":"https://pith.science/paper/HPUJUJ5L"},"agent_actions":{"view_html":"https://pith.science/pith/HPUJUJ5LJ7GPWUDTTZ4YLXWXC3","download_json":"https://pith.science/pith/HPUJUJ5LJ7GPWUDTTZ4YLXWXC3.json","view_paper":"https://pith.science/paper/HPUJUJ5L","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2207.02578&json=true","fetch_graph":"https://pith.science/api/pith-number/HPUJUJ5LJ7GPWUDTTZ4YLXWXC3/graph.json","fetch_events":"https://pith.science/api/pith-number/HPUJUJ5LJ7GPWUDTTZ4YLXWXC3/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/HPUJUJ5LJ7GPWUDTTZ4YLXWXC3/action/timestamp_anchor","attest_storage":"https://pith.science/pith/HPUJUJ5LJ7GPWUDTTZ4YLXWXC3/action/storage_attestation","attest_author":"https://pith.science/pith/HPUJUJ5LJ7GPWUDTTZ4YLXWXC3/action/author_attestation","sign_citation":"https://pith.science/pith/HPUJUJ5LJ7GPWUDTTZ4YLXWXC3/action/citation_signature","submit_replication":"https://pith.science/pith/HPUJUJ5LJ7GPWUDTTZ4YLXWXC3/action/replication_record"}},"created_at":"2026-07-05T06:09:31.389396+00:00","updated_at":"2026-07-05T06:09:31.389396+00:00"}