{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2020:DXS4WDPJIQJWOAGWJGFHEAQOQG","short_pith_number":"pith:DXS4WDPJ","schema_version":"1.0","canonical_sha256":"1de5cb0de944136700d6498a72020e81a2f3b96d6698b046efb377bcab03c0b7","source":{"kind":"arxiv","id":"2010.02650","version":2},"attestation_state":"computed","paper":{"title":"If beam search is the answer, what was the question?","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.CL","authors_text":"Clara Meister, Ryan Cotterell, Tim Vieira","submitted_at":"2020-10-06T11:57:03Z","abstract_excerpt":"Quite surprisingly, exact maximum a posteriori (MAP) decoding of neural language generators frequently leads to low-quality results. Rather, most state-of-the-art results on language generation tasks are attained using beam search despite its overwhelmingly high search error rate. This implies that the MAP objective alone does not express the properties we desire in text, which merits the question: if beam search is the answer, what was the question? We frame beam search as the exact solution to a different decoding objective in order to gain insights into why high probability under a model al"},"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":"2010.02650","kind":"arxiv","version":2},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2020-10-06T11:57:03Z","cross_cats_sorted":[],"title_canon_sha256":"b4f7f9cd29268e527f8f66c5e517bc880e6776c15ba61ad879bc4537ecab8c23","abstract_canon_sha256":"7ea60627c0ec9c756001c6870d8bd0231d5f41dd506b65de827ddb39ec847af2"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T02:07:31.272124Z","signature_b64":"5usQlgy4cLFcvAwqyqLZLn5oJQbkRiB6QfTcPH+gRL2aMtt3/kTcHoNHlccwnwzXGSb0s+m5nvKKyfaTRccxAA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"1de5cb0de944136700d6498a72020e81a2f3b96d6698b046efb377bcab03c0b7","last_reissued_at":"2026-07-05T02:07:31.271671Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T02:07:31.271671Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"If beam search is the answer, what was the question?","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.CL","authors_text":"Clara Meister, Ryan Cotterell, Tim Vieira","submitted_at":"2020-10-06T11:57:03Z","abstract_excerpt":"Quite surprisingly, exact maximum a posteriori (MAP) decoding of neural language generators frequently leads to low-quality results. Rather, most state-of-the-art results on language generation tasks are attained using beam search despite its overwhelmingly high search error rate. This implies that the MAP objective alone does not express the properties we desire in text, which merits the question: if beam search is the answer, what was the question? We frame beam search as the exact solution to a different decoding objective in order to gain insights into why high probability under a model al"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2010.02650","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/2010.02650/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":"2010.02650","created_at":"2026-07-05T02:07:31.271729+00:00"},{"alias_kind":"arxiv_version","alias_value":"2010.02650v2","created_at":"2026-07-05T02:07:31.271729+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2010.02650","created_at":"2026-07-05T02:07:31.271729+00:00"},{"alias_kind":"pith_short_12","alias_value":"DXS4WDPJIQJW","created_at":"2026-07-05T02:07:31.271729+00:00"},{"alias_kind":"pith_short_16","alias_value":"DXS4WDPJIQJWOAGW","created_at":"2026-07-05T02:07:31.271729+00:00"},{"alias_kind":"pith_short_8","alias_value":"DXS4WDPJ","created_at":"2026-07-05T02:07:31.271729+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":2,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2511.22277","citing_title":"TreeCoder: Systematic Exploration and Optimisation of Decoding and Constraints for LLM Code Generation","ref_index":33,"is_internal_anchor":false},{"citing_arxiv_id":"2604.23327","citing_title":"An Efficient Beam Search Algorithm for Active Perception in Mobile Robotics","ref_index":10,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/DXS4WDPJIQJWOAGWJGFHEAQOQG","json":"https://pith.science/pith/DXS4WDPJIQJWOAGWJGFHEAQOQG.json","graph_json":"https://pith.science/api/pith-number/DXS4WDPJIQJWOAGWJGFHEAQOQG/graph.json","events_json":"https://pith.science/api/pith-number/DXS4WDPJIQJWOAGWJGFHEAQOQG/events.json","paper":"https://pith.science/paper/DXS4WDPJ"},"agent_actions":{"view_html":"https://pith.science/pith/DXS4WDPJIQJWOAGWJGFHEAQOQG","download_json":"https://pith.science/pith/DXS4WDPJIQJWOAGWJGFHEAQOQG.json","view_paper":"https://pith.science/paper/DXS4WDPJ","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2010.02650&json=true","fetch_graph":"https://pith.science/api/pith-number/DXS4WDPJIQJWOAGWJGFHEAQOQG/graph.json","fetch_events":"https://pith.science/api/pith-number/DXS4WDPJIQJWOAGWJGFHEAQOQG/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/DXS4WDPJIQJWOAGWJGFHEAQOQG/action/timestamp_anchor","attest_storage":"https://pith.science/pith/DXS4WDPJIQJWOAGWJGFHEAQOQG/action/storage_attestation","attest_author":"https://pith.science/pith/DXS4WDPJIQJWOAGWJGFHEAQOQG/action/author_attestation","sign_citation":"https://pith.science/pith/DXS4WDPJIQJWOAGWJGFHEAQOQG/action/citation_signature","submit_replication":"https://pith.science/pith/DXS4WDPJIQJWOAGWJGFHEAQOQG/action/replication_record"}},"created_at":"2026-07-05T02:07:31.271729+00:00","updated_at":"2026-07-05T02:07:31.271729+00:00"}