{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2024:O3NJZI7Y25OVCTBVIFP6VFVTIJ","short_pith_number":"pith:O3NJZI7Y","schema_version":"1.0","canonical_sha256":"76da9ca3f8d75d514c35415fea96b3427c78c1dc5d1635ea69781cd39577ba0e","source":{"kind":"arxiv","id":"2406.17262","version":1},"attestation_state":"computed","paper":{"title":"D2LLM: Decomposed and Distilled Large Language Models for Semantic Search","license":"http://creativecommons.org/licenses/by-nc-nd/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.CL","authors_text":"Hang Yu, Jianguo Li, Jun Wang, Wei Zhang, Zihan Liao","submitted_at":"2024-06-25T04:03:04Z","abstract_excerpt":"The key challenge in semantic search is to create models that are both accurate and efficient in pinpointing relevant sentences for queries. While BERT-style bi-encoders excel in efficiency with pre-computed embeddings, they often miss subtle nuances in search tasks. Conversely, GPT-style LLMs with cross-encoder designs capture these nuances but are computationally intensive, hindering real-time applications. In this paper, we present D2LLMs-Decomposed and Distilled LLMs for semantic search-that combines the best of both worlds. We decompose a cross-encoder into an efficient bi-encoder integra"},"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":"2406.17262","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by-nc-nd/4.0/","primary_cat":"cs.CL","submitted_at":"2024-06-25T04:03:04Z","cross_cats_sorted":[],"title_canon_sha256":"4670c3450ab33876094ef57a4f8a1d2818157fff9f7cb9065f45df8538897174","abstract_canon_sha256":"03d48f769dfb176f7b4bc4ac513f332beef4b0df6fb3c764690050a9c5e2dd33"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T08:36:31.876842Z","signature_b64":"RxRtb43QYOHNL0zFTptW7AeECP6S3sEHsjWmBpjreMU+AHXU7HrJZ+64QdMf4P9CH1KT1UturUW6M0vraH1fDg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"76da9ca3f8d75d514c35415fea96b3427c78c1dc5d1635ea69781cd39577ba0e","last_reissued_at":"2026-07-05T08:36:31.876428Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T08:36:31.876428Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"D2LLM: Decomposed and Distilled Large Language Models for Semantic Search","license":"http://creativecommons.org/licenses/by-nc-nd/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.CL","authors_text":"Hang Yu, Jianguo Li, Jun Wang, Wei Zhang, Zihan Liao","submitted_at":"2024-06-25T04:03:04Z","abstract_excerpt":"The key challenge in semantic search is to create models that are both accurate and efficient in pinpointing relevant sentences for queries. While BERT-style bi-encoders excel in efficiency with pre-computed embeddings, they often miss subtle nuances in search tasks. Conversely, GPT-style LLMs with cross-encoder designs capture these nuances but are computationally intensive, hindering real-time applications. In this paper, we present D2LLMs-Decomposed and Distilled LLMs for semantic search-that combines the best of both worlds. We decompose a cross-encoder into an efficient bi-encoder integra"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2406.17262","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/2406.17262/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":"2406.17262","created_at":"2026-07-05T08:36:31.876483+00:00"},{"alias_kind":"arxiv_version","alias_value":"2406.17262v1","created_at":"2026-07-05T08:36:31.876483+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2406.17262","created_at":"2026-07-05T08:36:31.876483+00:00"},{"alias_kind":"pith_short_12","alias_value":"O3NJZI7Y25OV","created_at":"2026-07-05T08:36:31.876483+00:00"},{"alias_kind":"pith_short_16","alias_value":"O3NJZI7Y25OVCTBV","created_at":"2026-07-05T08:36:31.876483+00:00"},{"alias_kind":"pith_short_8","alias_value":"O3NJZI7Y","created_at":"2026-07-05T08:36:31.876483+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":0,"internal_anchor_count":0,"sample":[]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/O3NJZI7Y25OVCTBVIFP6VFVTIJ","json":"https://pith.science/pith/O3NJZI7Y25OVCTBVIFP6VFVTIJ.json","graph_json":"https://pith.science/api/pith-number/O3NJZI7Y25OVCTBVIFP6VFVTIJ/graph.json","events_json":"https://pith.science/api/pith-number/O3NJZI7Y25OVCTBVIFP6VFVTIJ/events.json","paper":"https://pith.science/paper/O3NJZI7Y"},"agent_actions":{"view_html":"https://pith.science/pith/O3NJZI7Y25OVCTBVIFP6VFVTIJ","download_json":"https://pith.science/pith/O3NJZI7Y25OVCTBVIFP6VFVTIJ.json","view_paper":"https://pith.science/paper/O3NJZI7Y","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2406.17262&json=true","fetch_graph":"https://pith.science/api/pith-number/O3NJZI7Y25OVCTBVIFP6VFVTIJ/graph.json","fetch_events":"https://pith.science/api/pith-number/O3NJZI7Y25OVCTBVIFP6VFVTIJ/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/O3NJZI7Y25OVCTBVIFP6VFVTIJ/action/timestamp_anchor","attest_storage":"https://pith.science/pith/O3NJZI7Y25OVCTBVIFP6VFVTIJ/action/storage_attestation","attest_author":"https://pith.science/pith/O3NJZI7Y25OVCTBVIFP6VFVTIJ/action/author_attestation","sign_citation":"https://pith.science/pith/O3NJZI7Y25OVCTBVIFP6VFVTIJ/action/citation_signature","submit_replication":"https://pith.science/pith/O3NJZI7Y25OVCTBVIFP6VFVTIJ/action/replication_record"}},"created_at":"2026-07-05T08:36:31.876483+00:00","updated_at":"2026-07-05T08:36:31.876483+00:00"}