{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2022:BS67I6TOV7BFW7DY6RAYVEJ6G2","short_pith_number":"pith:BS67I6TO","schema_version":"1.0","canonical_sha256":"0cbdf47a6eafc25b7c78f4418a913e3695e5077d83ee737c4c5ec10c156032ff","source":{"kind":"arxiv","id":"2202.08904","version":5},"attestation_state":"computed","paper":{"title":"SGPT: GPT Sentence Embeddings for Semantic Search","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.AI","cs.IR"],"primary_cat":"cs.CL","authors_text":"Niklas Muennighoff","submitted_at":"2022-02-17T21:35:56Z","abstract_excerpt":"Decoder transformers have continued increasing in scale reaching hundreds of billions of parameters. Due to their scale the same decoder sets state-of-the-art results on various language tasks via prompting or fine-tuning. Yet, these large foundation models remain unusable for the related fields of semantic search and sentence embeddings. This prevents possibly new state-of-the-art results and forces organizations to train and maintain separate models. To this end, we propose SGPT to use decoders for sentence embeddings and semantic search via prompting or fine-tuning. At 5.8 billion parameter"},"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":"2202.08904","kind":"arxiv","version":5},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2022-02-17T21:35:56Z","cross_cats_sorted":["cs.AI","cs.IR"],"title_canon_sha256":"c008fcf9d14b83b5ab5c32da510bb5c8eddb0875231472eeb426346a4cfa427c","abstract_canon_sha256":"df51c64d7aad8984dea175173bd4d2efdb9eeb597907e84cf79a6bbda7750cce"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T04:46:13.618774Z","signature_b64":"qGSXhhCXtN5Tnwt2czleDMyUj+8U2n/CVLW47F11uT/Kgb5EFtk0ig5eaDy4SiSh0dlM/lW/hoTZIR5oAohwCg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"0cbdf47a6eafc25b7c78f4418a913e3695e5077d83ee737c4c5ec10c156032ff","last_reissued_at":"2026-07-05T04:46:13.618305Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T04:46:13.618305Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"SGPT: GPT Sentence Embeddings for Semantic Search","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.AI","cs.IR"],"primary_cat":"cs.CL","authors_text":"Niklas Muennighoff","submitted_at":"2022-02-17T21:35:56Z","abstract_excerpt":"Decoder transformers have continued increasing in scale reaching hundreds of billions of parameters. Due to their scale the same decoder sets state-of-the-art results on various language tasks via prompting or fine-tuning. Yet, these large foundation models remain unusable for the related fields of semantic search and sentence embeddings. This prevents possibly new state-of-the-art results and forces organizations to train and maintain separate models. To this end, we propose SGPT to use decoders for sentence embeddings and semantic search via prompting or fine-tuning. At 5.8 billion parameter"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2202.08904","kind":"arxiv","version":5},"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/2202.08904/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":"2202.08904","created_at":"2026-07-05T04:46:13.618376+00:00"},{"alias_kind":"arxiv_version","alias_value":"2202.08904v5","created_at":"2026-07-05T04:46:13.618376+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2202.08904","created_at":"2026-07-05T04:46:13.618376+00:00"},{"alias_kind":"pith_short_12","alias_value":"BS67I6TOV7BF","created_at":"2026-07-05T04:46:13.618376+00:00"},{"alias_kind":"pith_short_16","alias_value":"BS67I6TOV7BFW7DY","created_at":"2026-07-05T04:46:13.618376+00:00"},{"alias_kind":"pith_short_8","alias_value":"BS67I6TO","created_at":"2026-07-05T04:46:13.618376+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":23,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2606.25674","citing_title":"BitNet Text Embeddings","ref_index":47,"is_internal_anchor":false},{"citing_arxiv_id":"2606.05858","citing_title":"ReverseEOL: Improving Training-free Text Embeddings via Text Reversal in Decoder-only LLMs","ref_index":45,"is_internal_anchor":false},{"citing_arxiv_id":"2606.00203","citing_title":"DeSQ: Decomposition-based SPARQL Query Generation","ref_index":37,"is_internal_anchor":false},{"citing_arxiv_id":"2401.03563","citing_title":"Data-CUBE: Data Curriculum for Instruction-based Sentence Representation Learning","ref_index":32,"is_internal_anchor":false},{"citing_arxiv_id":"2404.10981","citing_title":"A Survey on Retrieval-Augmented Text Generation for Large Language Models","ref_index":102,"is_internal_anchor":false},{"citing_arxiv_id":"2605.18766","citing_title":"Retrieve Only Relevant Tables Whether Few or Many: Adaptive Table Retrieval Method","ref_index":55,"is_internal_anchor":false},{"citing_arxiv_id":"2507.00994","citing_title":"Should We Still Pretrain Encoders with Masked Language Modeling?","ref_index":29,"is_internal_anchor":false},{"citing_arxiv_id":"2507.23386","citing_title":"Causal2Vec: Improving Decoder-only LLMs as Embedding Models through a Contextual Token","ref_index":31,"is_internal_anchor":false},{"citing_arxiv_id":"2510.07048","citing_title":"Search-R3: Unifying Reasoning and Embedding in Large Language Models","ref_index":48,"is_internal_anchor":false},{"citing_arxiv_id":"2305.16264","citing_title":"Scaling Data-Constrained Language Models","ref_index":77,"is_internal_anchor":false},{"citing_arxiv_id":"2301.12652","citing_title":"REPLUG: Retrieval-Augmented Black-Box Language Models","ref_index":210,"is_internal_anchor":false},{"citing_arxiv_id":"2407.12580","citing_title":"E5-V: Universal Embeddings with Multimodal Large Language Models","ref_index":9,"is_internal_anchor":false},{"citing_arxiv_id":"2410.10594","citing_title":"VisRAG: Vision-based Retrieval-augmented Generation on Multi-modality Documents","ref_index":15,"is_internal_anchor":false},{"citing_arxiv_id":"2410.05229","citing_title":"GSM-Symbolic: Understanding the Limitations of Mathematical Reasoning in Large Language Models","ref_index":90,"is_internal_anchor":false},{"citing_arxiv_id":"2604.23336","citing_title":"Efficient Rationale-based Retrieval: On-policy Distillation from Generative Rerankers based on JEPA","ref_index":16,"is_internal_anchor":false},{"citing_arxiv_id":"2309.07597","citing_title":"C-Pack: Packed Resources For General Chinese Embeddings","ref_index":35,"is_internal_anchor":false},{"citing_arxiv_id":"2211.05100","citing_title":"BLOOM: A 176B-Parameter Open-Access Multilingual Language Model","ref_index":285,"is_internal_anchor":false},{"citing_arxiv_id":"2402.03216","citing_title":"M3-Embedding: Multi-Linguality, Multi-Functionality, Multi-Granularity Text Embeddings Through Self-Knowledge Distillation","ref_index":57,"is_internal_anchor":false},{"citing_arxiv_id":"2604.23336","citing_title":"Efficient Rationale-based Retrieval: On-policy Distillation from Generative Rerankers based on JEPA","ref_index":16,"is_internal_anchor":false},{"citing_arxiv_id":"2605.04962","citing_title":"TabEmbed: Benchmarking and Learning Generalist Embeddings for Tabular Understanding","ref_index":22,"is_internal_anchor":false},{"citing_arxiv_id":"2605.01372","citing_title":"Embedding-based In-Context Prompt Training for Enhancing LLMs as Text Encoders","ref_index":60,"is_internal_anchor":false},{"citing_arxiv_id":"2212.03533","citing_title":"Text Embeddings by Weakly-Supervised Contrastive Pre-training","ref_index":41,"is_internal_anchor":false},{"citing_arxiv_id":"2604.15591","citing_title":"BioHiCL: Hierarchical Multi-Label Contrastive Learning for Biomedical Retrieval with MeSH Labels","ref_index":3,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/BS67I6TOV7BFW7DY6RAYVEJ6G2","json":"https://pith.science/pith/BS67I6TOV7BFW7DY6RAYVEJ6G2.json","graph_json":"https://pith.science/api/pith-number/BS67I6TOV7BFW7DY6RAYVEJ6G2/graph.json","events_json":"https://pith.science/api/pith-number/BS67I6TOV7BFW7DY6RAYVEJ6G2/events.json","paper":"https://pith.science/paper/BS67I6TO"},"agent_actions":{"view_html":"https://pith.science/pith/BS67I6TOV7BFW7DY6RAYVEJ6G2","download_json":"https://pith.science/pith/BS67I6TOV7BFW7DY6RAYVEJ6G2.json","view_paper":"https://pith.science/paper/BS67I6TO","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2202.08904&json=true","fetch_graph":"https://pith.science/api/pith-number/BS67I6TOV7BFW7DY6RAYVEJ6G2/graph.json","fetch_events":"https://pith.science/api/pith-number/BS67I6TOV7BFW7DY6RAYVEJ6G2/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/BS67I6TOV7BFW7DY6RAYVEJ6G2/action/timestamp_anchor","attest_storage":"https://pith.science/pith/BS67I6TOV7BFW7DY6RAYVEJ6G2/action/storage_attestation","attest_author":"https://pith.science/pith/BS67I6TOV7BFW7DY6RAYVEJ6G2/action/author_attestation","sign_citation":"https://pith.science/pith/BS67I6TOV7BFW7DY6RAYVEJ6G2/action/citation_signature","submit_replication":"https://pith.science/pith/BS67I6TOV7BFW7DY6RAYVEJ6G2/action/replication_record"}},"created_at":"2026-07-05T04:46:13.618376+00:00","updated_at":"2026-07-05T04:46:13.618376+00:00"}