{"id":"https://openalex.org/W4394591144","doi":"https://doi.org/10.48550/arxiv.2404.03683","title":"Stream of Search (SoS): Learning to Search in Language","display_name":"Stream of Search (SoS): Learning to Search in Language","publication_year":2024,"publication_date":"2024-04-01","ids":{"openalex":"https://openalex.org/W4394591144","doi":"https://doi.org/10.48550/arxiv.2404.03683"},"language":"en","primary_location":{"id":"pmh:oai:arXiv.org:2404.03683","doi":null,"is_oa":true,"landing_page_url":"http://arxiv.org/abs/2404.03683","pdf_url":"https://arxiv.org/pdf/2404.03683","source":{"id":"https://openalex.org/S4306400194","display_name":"arXiv (Cornell University)","issn_l":"2331-8422","issn":["2331-8422"],"is_oa":true,"is_in_doaj":false,"is_core":false,"listed_in":[],"host_organization":"https://openalex.org/I205783295","host_organization_name":"Cornell University","host_organization_lineage":["https://openalex.org/I205783295"],"host_organization_lineage_names":[],"type":"repository"},"license":"cc-by","license_id":"https://openalex.org/licenses/cc-by","version":"submittedVersion","is_accepted":false,"is_published":false,"raw_source_name":null,"raw_type":"text"},"volume":null,"type":"preprint","indexed_in":["arxiv","datacite"],"open_access":{"is_oa":true,"oa_status":"green","oa_url":"https://arxiv.org/pdf/2404.03683","any_repository_has_fulltext":true},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5085625870","display_name":"Kanishk Gandhi","orcid":null,"observed_orcids":[]},"institutions":[{"id":"https://openalex.org/I97018004","display_name":"Stanford University","ror":"https://ror.org/00f54p054","country_code":"US","type":"education","lineage":["https://openalex.org/I97018004"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Gandhi, Kanishk","raw_affiliation_strings":["Stanford University"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Stanford University","institution_ids":["https://openalex.org/I97018004"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5114406738","display_name":"Denise T Lee","orcid":"https://orcid.org/0009-0000-9203-925X","observed_orcids":["https://orcid.org/0009-0000-9203-925X"]},"institutions":[{"id":"https://openalex.org/I97018004","display_name":"Stanford University","ror":"https://ror.org/00f54p054","country_code":"US","type":"education","lineage":["https://openalex.org/I97018004"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Lee, Denise","raw_affiliation_strings":["Stanford University"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Stanford University","institution_ids":["https://openalex.org/I97018004"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5082051197","display_name":"Gabriel Grand","orcid":"https://orcid.org/0000-0003-1920-0021","observed_orcids":["https://orcid.org/0000-0003-1920-0021"]},"institutions":[{"id":"https://openalex.org/I63966007","display_name":"Massachusetts Institute of Technology","ror":"https://ror.org/042nb2s44","country_code":"US","type":"education","lineage":["https://openalex.org/I63966007"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Grand, Gabriel","raw_affiliation_strings":["MIT"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"MIT","institution_ids":["https://openalex.org/I63966007"]}]},{"author_position":"middle","author":{"id":null,"display_name":"Liu, Muxin","orcid":null,"observed_orcids":[]},"institutions":[{"id":"https://openalex.org/I63966007","display_name":"Massachusetts Institute of Technology","ror":"https://ror.org/042nb2s44","country_code":"US","type":"education","lineage":["https://openalex.org/I63966007"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Liu, Muxin","raw_affiliation_strings":["MIT"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"MIT","institution_ids":["https://openalex.org/I63966007"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5084815985","display_name":"Winson Cheng","orcid":null,"observed_orcids":[]},"institutions":[{"id":"https://openalex.org/I97018004","display_name":"Stanford University","ror":"https://ror.org/00f54p054","country_code":"US","type":"education","lineage":["https://openalex.org/I97018004"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Cheng, Winson","raw_affiliation_strings":["Stanford University"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Stanford University","institution_ids":["https://openalex.org/I97018004"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5127868678","display_name":"Archit Sharma","orcid":null,"observed_orcids":[]},"institutions":[{"id":"https://openalex.org/I97018004","display_name":"Stanford University","ror":"https://ror.org/00f54p054","country_code":"US","type":"education","lineage":["https://openalex.org/I97018004"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Sharma, Archit","raw_affiliation_strings":["Stanford University"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Stanford University","institution_ids":["https://openalex.org/I97018004"]}]},{"author_position":"last","author":{"id":"https://openalex.org/A5001961716","display_name":"Noah D. Goodman","orcid":"https://orcid.org/0000-0002-9176-8802","observed_orcids":["https://orcid.org/0000-0002-9176-8802"]},"institutions":[{"id":"https://openalex.org/I97018004","display_name":"Stanford University","ror":"https://ror.org/00f54p054","country_code":"US","type":"education","lineage":["https://openalex.org/I97018004"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Goodman, Noah D.","raw_affiliation_strings":["Stanford University"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Stanford University","institution_ids":["https://openalex.org/I97018004"]}]}],"institutions":[],"countries_distinct_count":1,"institutions_distinct_count":2,"corresponding_author_ids":[],"corresponding_institution_ids":[],"apc_list":null,"apc_paid":null,"fwci":null,"has_fulltext":true,"cited_by_count":2,"citation_normalized_percentile":null,"cited_by_percentile_year":null,"biblio":{"volume":null,"issue":null,"first_page":null,"last_page":null},"is_retracted":false,"is_paratext":false,"is_xpac":false,"primary_topic":{"id":"https://openalex.org/T10906","display_name":"AI-based Problem Solving and Planning","score":0.37869998812675476,"subfield":{"id":"https://openalex.org/subfields/1702","display_name":"Artificial Intelligence"},"field":{"id":"https://openalex.org/fields/17","display_name":"Computer Science"},"domain":{"id":"https://openalex.org/domains/3","display_name":"Physical Sciences"}},"topics":[{"id":"https://openalex.org/T10906","display_name":"AI-based Problem Solving and Planning","score":0.37869998812675476,"subfield":{"id":"https://openalex.org/subfields/1702","display_name":"Artificial Intelligence"},"field":{"id":"https://openalex.org/fields/17","display_name":"Computer Science"},"domain":{"id":"https://openalex.org/domains/3","display_name":"Physical Sciences"}},{"id":"https://openalex.org/T10028","display_name":"Topic Modeling","score":0.3075999915599823,"subfield":{"id":"https://openalex.org/subfields/1702","display_name":"Artificial Intelligence"},"field":{"id":"https://openalex.org/fields/17","display_name":"Computer Science"},"domain":{"id":"https://openalex.org/domains/3","display_name":"Physical Sciences"}},{"id":"https://openalex.org/T10181","display_name":"Natural Language Processing Techniques","score":0.08910000324249268,"subfield":{"id":"https://openalex.org/subfields/1702","display_name":"Artificial Intelligence"},"field":{"id":"https://openalex.org/fields/17","display_name":"Computer Science"},"domain":{"id":"https://openalex.org/domains/3","display_name":"Physical Sciences"}}],"keywords":[{"id":"https://openalex.org/keywords/arithmetic-problem-solving","display_name":"arithmetic problem solving","score":0.44699999690055847},{"id":"https://openalex.org/keywords/heuristic-solver","display_name":"heuristic solver","score":0.6510000228881836},{"id":"https://openalex.org/keywords/transformer-pre-training","display_name":"transformer pre-training","score":0.628000020980835}],"concepts":[{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.5384007096290588},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.3651059865951538},{"id":"https://openalex.org/C204321447","wikidata":"https://www.wikidata.org/wiki/Q30642","display_name":"Natural language processing","level":1,"score":0.35532453656196594}],"mesh":[],"locations_count":2,"locations":[{"id":"pmh:oai:arXiv.org:2404.03683","doi":null,"is_oa":true,"landing_page_url":"http://arxiv.org/abs/2404.03683","pdf_url":"https://arxiv.org/pdf/2404.03683","source":{"id":"https://openalex.org/S4306400194","display_name":"arXiv (Cornell University)","issn_l":"2331-8422","issn":["2331-8422"],"is_oa":true,"is_in_doaj":false,"is_core":false,"listed_in":[],"host_organization":"https://openalex.org/I205783295","host_organization_name":"Cornell University","host_organization_lineage":["https://openalex.org/I205783295"],"host_organization_lineage_names":[],"type":"repository"},"license":"cc-by","license_id":"https://openalex.org/licenses/cc-by","version":"submittedVersion","is_accepted":false,"is_published":false,"raw_source_name":null,"raw_type":"text"},{"id":"doi:10.48550/arxiv.2404.03683","doi":"https://doi.org/10.48550/arxiv.2404.03683","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2404.03683","pdf_url":null,"source":{"id":"https://openalex.org/S4306400194","display_name":"arXiv (Cornell University)","issn_l":"2331-8422","issn":["2331-8422"],"is_oa":true,"is_in_doaj":false,"is_core":false,"listed_in":[],"host_organization":"https://openalex.org/I205783295","host_organization_name":"Cornell University","host_organization_lineage":["https://openalex.org/I205783295"],"host_organization_lineage_names":[],"type":"repository"},"license":"cc-by","license_id":"https://openalex.org/licenses/cc-by","version":null,"is_accepted":false,"is_published":false,"raw_source_name":null,"raw_type":"Preprint"}],"best_oa_location":{"id":"pmh:oai:arXiv.org:2404.03683","doi":null,"is_oa":true,"landing_page_url":"http://arxiv.org/abs/2404.03683","pdf_url":"https://arxiv.org/pdf/2404.03683","source":{"id":"https://openalex.org/S4306400194","display_name":"arXiv (Cornell University)","issn_l":"2331-8422","issn":["2331-8422"],"is_oa":true,"is_in_doaj":false,"is_core":false,"listed_in":[],"host_organization":"https://openalex.org/I205783295","host_organization_name":"Cornell University","host_organization_lineage":["https://openalex.org/I205783295"],"host_organization_lineage_names":[],"type":"repository"},"license":"cc-by","license_id":"https://openalex.org/licenses/cc-by","version":"submittedVersion","is_accepted":false,"is_published":false,"raw_source_name":null,"raw_type":"text"},"sustainable_development_goals":[],"study_designs":[],"awards":[{"id":"https://openalex.org/G1657816209","display_name":"Expeditions: Collaborative Research: Understanding the World Through Code","funder_award_id":"1918771","funder_id":"https://openalex.org/F4320306076","funder_display_name":"National Science Foundation"}],"funders":[{"id":"https://openalex.org/F4320306076","display_name":"National Science Foundation","ror":"https://ror.org/021nxhr62"}],"has_content":{"grobid_xml":true,"pdf":true},"content_urls":{"pdf":"https://content.openalex.org/works/W4394591144.pdf","grobid_xml":"https://content.openalex.org/works/W4394591144.grobid-xml"},"referenced_works_count":0,"referenced_works":[],"related_works":["https://openalex.org/W7197033118","https://openalex.org/W7162634757","https://openalex.org/W4416877088","https://openalex.org/W7202290415","https://openalex.org/W4412887031","https://openalex.org/W7141165457","https://openalex.org/W7133209886","https://openalex.org/W7167698968","https://openalex.org/W4415190582","https://openalex.org/W4415335861"],"abstract_inverted_index":{"Language":[0],"models":[1,43,141,171,195],"are":[2],"rarely":[3],"shown":[4],"fruitful":[5],"mistakes":[6],"while":[7],"training.":[8],"They":[9],"then":[10],"struggle":[11],"to":[12,26,47,99,106,143,198,204],"look":[13],"beyond":[14],"the":[15,28,51,88,96,146,187],"next":[16],"token,":[17],"suffering":[18],"from":[19,117],"a":[20,58,62,69,108,113,120],"snowballing":[21],"of":[22,30,53,64,78,93,122,124,174,186],"errors":[23],"and":[24,164,210],"struggling":[25],"predict":[27,144],"consequence":[29],"their":[31],"actions":[32],"several":[33],"steps":[34],"ahead.":[35],"In":[36],"this":[37,153],"paper,":[38],"we":[39],"show":[40],"how":[41],"language":[42,71,115,194],"can":[44,196],"be":[45,182],"taught":[46],"search":[48,54,65,73,81,125,136,148,208],"by":[49,127,138,184],"representing":[50],"process":[52],"in":[55],"language,":[56],"as":[57],"flattened":[59],"string":[60],"--":[61],"stream":[63],"(SoS).":[66],"We":[67,83,111,130,150],"propose":[68],"unified":[70],"for":[72],"that":[74,132,180,193],"captures":[75],"an":[76],"array":[77],"different":[79,207],"symbolic":[80],"strategies.":[82],"demonstrate":[84],"our":[85],"approach":[86],"using":[87],"simple":[89],"yet":[90],"difficult":[91],"game":[92],"Countdown,":[94],"where":[95],"goal":[97],"is":[98],"combine":[100],"input":[101],"numbers":[102],"with":[103,155],"arithmetic":[104],"operations":[105],"reach":[107],"target":[109],"number.":[110],"pretrain":[112],"transformer-based":[114],"model":[116,154],"scratch":[118],"on":[119],"dataset":[121],"streams":[123],"generated":[126],"heuristic":[128,188],"solvers.":[129,189],"find":[131],"SoS":[133,170],"pretraining":[134],"increases":[135],"accuracy":[137],"25%":[139],"over":[140],"trained":[142],"only":[145],"optimal":[147],"trajectory.":[149],"further":[151],"finetune":[152],"two":[156],"policy":[157],"improvement":[158],"methods:":[159],"Advantage-Induced":[160],"Policy":[161],"Alignment":[162],"(APA)":[163],"Self-Taught":[165],"Reasoner":[166],"(STaR).":[167],"The":[168],"finetuned":[169],"solve":[172,199],"36%":[173],"previously":[175],"unsolved":[176],"problems,":[177],"including":[178],"problems":[179,200],"cannot":[181],"solved":[183],"any":[185],"Our":[190],"results":[191],"indicate":[192],"learn":[197],"via":[201],"search,":[202],"self-improve":[203],"flexibly":[205],"use":[206],"strategies,":[209],"potentially":[211],"discover":[212],"new":[213],"ones.":[214]},"counts_by_year":[{"year":2025,"cited_by_count":2}],"updated_date":"2026-10-14T16:39:01.820549","created_date":"2024-04-09T00:00:00"}