{"id":"https://openalex.org/W7163357448","doi":"https://doi.org/10.48550/arxiv.2606.03782","title":"Reasoning over Grammar: Can Synthetic Linguistic Reasoning Traces Enhance Low-Resource Machine Translation?","display_name":"Reasoning over Grammar: Can Synthetic Linguistic Reasoning Traces Enhance Low-Resource Machine Translation?","publication_year":2026,"publication_date":"2026-06-02","ids":{"openalex":"https://openalex.org/W7163357448","doi":"https://doi.org/10.48550/arxiv.2606.03782"},"language":null,"primary_location":{"id":"doi:10.48550/arxiv.2606.03782","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2606.03782","pdf_url":null,"source":{"id":"https://openalex.org/S4306400194","display_name":"arXiv (Cornell University)","issn_l":null,"issn":null,"is_oa":true,"is_in_doaj":false,"is_core":false,"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":null,"license_id":null,"version":null,"is_accepted":false,"is_published":false,"raw_source_name":null,"raw_type":"Preprint"},"type":"preprint","indexed_in":["datacite"],"open_access":{"is_oa":true,"oa_status":"green","oa_url":"https://doi.org/10.48550/arxiv.2606.03782","any_repository_has_fulltext":true},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5018686674","display_name":"Renhao Pei","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Pei, Renhao","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5137735266","display_name":"Yihong Liu","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Liu, Yihong","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5137738567","display_name":"Sampo Pyysalo","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Pyysalo, Sampo","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5137788083","display_name":"Hinrich Sch\u00fctze","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Sch\u00fctze, Hinrich","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"last","author":{"id":"https://openalex.org/A5007841686","display_name":"Shaoxiong Ji","orcid":"https://orcid.org/0000-0003-3281-8002"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Ji, Shaoxiong","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]}],"institutions":[],"countries_distinct_count":0,"institutions_distinct_count":0,"corresponding_author_ids":[],"corresponding_institution_ids":[],"apc_list":null,"apc_paid":null,"fwci":null,"has_fulltext":false,"cited_by_count":0,"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/T10181","display_name":"Natural Language Processing Techniques","score":0.7759000062942505,"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/T10181","display_name":"Natural Language Processing Techniques","score":0.7759000062942505,"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":0.06780000030994415,"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/T11714","display_name":"Multimodal Machine Learning Applications","score":0.017100000753998756,"subfield":{"id":"https://openalex.org/subfields/1707","display_name":"Computer Vision and Pattern Recognition"},"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/leverage","display_name":"Leverage (statistics)","score":0.583899974822998},{"id":"https://openalex.org/keywords/deep-linguistic-processing","display_name":"Deep linguistic processing","score":0.5432999730110168},{"id":"https://openalex.org/keywords/rule-based-machine-translation","display_name":"Rule-based machine translation","score":0.5078999996185303},{"id":"https://openalex.org/keywords/pipeline","display_name":"Pipeline (software)","score":0.4544999897480011},{"id":"https://openalex.org/keywords/trace","display_name":"TRACE (psycholinguistics)","score":0.4503999948501587},{"id":"https://openalex.org/keywords/linguistic-description","display_name":"Linguistic description","score":0.400299996137619},{"id":"https://openalex.org/keywords/machine-translation","display_name":"Machine translation","score":0.3508000075817108}],"concepts":[{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.6685000061988831},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.609499990940094},{"id":"https://openalex.org/C153083717","wikidata":"https://www.wikidata.org/wiki/Q6535263","display_name":"Leverage (statistics)","level":2,"score":0.583899974822998},{"id":"https://openalex.org/C64249296","wikidata":"https://www.wikidata.org/wiki/Q5250411","display_name":"Deep linguistic processing","level":2,"score":0.5432999730110168},{"id":"https://openalex.org/C204321447","wikidata":"https://www.wikidata.org/wiki/Q30642","display_name":"Natural language processing","level":1,"score":0.536300003528595},{"id":"https://openalex.org/C53893814","wikidata":"https://www.wikidata.org/wiki/Q7378909","display_name":"Rule-based machine translation","level":2,"score":0.5078999996185303},{"id":"https://openalex.org/C43521106","wikidata":"https://www.wikidata.org/wiki/Q2165493","display_name":"Pipeline (software)","level":2,"score":0.4544999897480011},{"id":"https://openalex.org/C75291252","wikidata":"https://www.wikidata.org/wiki/Q1315756","display_name":"TRACE (psycholinguistics)","level":2,"score":0.4503999948501587},{"id":"https://openalex.org/C172205122","wikidata":"https://www.wikidata.org/wiki/Q777864","display_name":"Linguistic description","level":2,"score":0.400299996137619},{"id":"https://openalex.org/C41895202","wikidata":"https://www.wikidata.org/wiki/Q8162","display_name":"Linguistics","level":1,"score":0.3824000060558319},{"id":"https://openalex.org/C203005215","wikidata":"https://www.wikidata.org/wiki/Q79798","display_name":"Machine translation","level":2,"score":0.3508000075817108},{"id":"https://openalex.org/C206977991","wikidata":"https://www.wikidata.org/wiki/Q980357","display_name":"Grammatical category","level":3,"score":0.3269999921321869},{"id":"https://openalex.org/C2987219923","wikidata":"https://www.wikidata.org/wiki/Q777864","display_name":"Linguistic analysis","level":2,"score":0.3125},{"id":"https://openalex.org/C155092808","wikidata":"https://www.wikidata.org/wiki/Q182557","display_name":"Computational linguistics","level":2,"score":0.311599999666214},{"id":"https://openalex.org/C193221554","wikidata":"https://www.wikidata.org/wiki/Q5153664","display_name":"Commonsense reasoning","level":2,"score":0.29019999504089355},{"id":"https://openalex.org/C149364088","wikidata":"https://www.wikidata.org/wiki/Q185917","display_name":"Translation (biology)","level":4,"score":0.2867000102996826},{"id":"https://openalex.org/C44291984","wikidata":"https://www.wikidata.org/wiki/Q1074173","display_name":"Question answering","level":2,"score":0.28369998931884766},{"id":"https://openalex.org/C195344581","wikidata":"https://www.wikidata.org/wiki/Q2555318","display_name":"Automated reasoning","level":2,"score":0.2770000100135803},{"id":"https://openalex.org/C194926167","wikidata":"https://www.wikidata.org/wiki/Q6554066","display_name":"Linguistic sequence complexity","level":2,"score":0.2653999924659729}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.48550/arxiv.2606.03782","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2606.03782","pdf_url":null,"source":{"id":"https://openalex.org/S4306400194","display_name":"arXiv (Cornell University)","issn_l":null,"issn":null,"is_oa":true,"is_in_doaj":false,"is_core":false,"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":null,"license_id":null,"version":null,"is_accepted":false,"is_published":null,"raw_source_name":null,"raw_type":"Preprint"}],"best_oa_location":{"id":"doi:10.48550/arxiv.2606.03782","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2606.03782","pdf_url":null,"source":{"id":"https://openalex.org/S4306400194","display_name":"arXiv (Cornell University)","issn_l":null,"issn":null,"is_oa":true,"is_in_doaj":false,"is_core":false,"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":null,"license_id":null,"version":null,"is_accepted":false,"is_published":false,"raw_source_name":null,"raw_type":"Preprint"},"sustainable_development_goals":[{"id":"https://metadata.un.org/sdg/4","display_name":"Quality Education","score":0.6528401374816895}],"awards":[],"funders":[],"has_content":{"pdf":false,"grobid_xml":false},"content_urls":null,"referenced_works_count":0,"referenced_works":[],"related_works":[],"abstract_inverted_index":{"Large":[0],"language":[1],"models":[2,146],"(LLMs)":[3],"offer":[4],"a":[5,60,180],"promising":[6],"approach":[7],"to":[8,27,175],"machine":[9],"translation":[10,121],"(MT)":[11],"for":[12,62,165],"extremely":[13],"low-resource":[14,44,166],"languages":[15],"by":[16,35],"incorporating":[17],"linguistic":[18,53,66,105,133,171],"resources":[19],"through":[20],"in-context":[21,84],"learning.":[22],"However,":[23],"LLMs":[24,160],"often":[25,152],"struggle":[26],"apply":[28],"grammatical":[29,56,163],"information":[30,164],"effectively":[31],"during":[32],"translation.":[33],"Inspired":[34],"recent":[36],"progress":[37],"in":[38,81,114],"chain-of-thought":[39],"reasoning,":[40],"we":[41],"investigate":[42],"whether":[43],"MT":[45,167],"can":[46,161],"benefit":[47],"from":[48,69],"structured":[49],"intermediate":[50],"steps":[51],"of":[52],"analysis":[54],"and":[55,74,90,96,127,141],"reasoning.":[57],"We":[58,77],"propose":[59],"pipeline":[61],"automatically":[63],"generating":[64],"step-by-step":[65],"reasoning":[67,106,134],"traces":[68,80,107,118,135],"Universal":[70],"Dependencies":[71],"treebanks,":[72],"dictionaries,":[73],"grammar-rule":[75],"banks.":[76],"evaluate":[78],"these":[79],"three":[82],"settings:":[83],"learning":[85,174],"(ICL),":[86],"supervised":[87],"fine-tuning":[88,92],"(SFT),":[89],"reinforcement":[91],"(RFT),":[93],"on":[94],"Xibe":[95],"Chintang":[97],"as":[98,111,136,145],"test":[99],"cases.":[100],"Our":[101],"results":[102],"show":[103],"that":[104,159],"are":[108],"most":[109,124],"effective":[110],"inference-time":[112],"guidance:":[113],"ICL,":[115],"reliable":[116,170],"sentence-specific":[117],"substantially":[119],"improve":[120],"performance":[122],"across":[123],"models,":[125],"languages,":[126],"metrics.":[128],"In":[129],"contrast,":[130],"using":[131],"the":[132,148],"training":[137],"data":[138],"yields":[139],"smaller":[140],"less":[142],"consistent":[143],"gains,":[144],"learn":[147],"trace":[149],"format":[150],"but":[151],"generate":[153,176],"erroneous":[154],"content.":[155],"These":[156],"findings":[157],"suggest":[158],"leverage":[162],"when":[168],"given":[169],"analyses,":[172],"while":[173],"such":[177],"analyses":[178],"remains":[179],"major":[181],"bottleneck.":[182]},"counts_by_year":[],"updated_date":"2026-07-01T06:00:48.157686","created_date":"2026-06-04T00:00:00"}