{"id":"https://openalex.org/W7162393354","doi":"https://doi.org/10.48550/arxiv.2605.24869","title":"Lngram: N-gram Conditional Memory in Latent Space","display_name":"Lngram: N-gram Conditional Memory in Latent Space","publication_year":2026,"publication_date":"2026-05-24","ids":{"openalex":"https://openalex.org/W7162393354","doi":"https://doi.org/10.48550/arxiv.2605.24869"},"language":null,"primary_location":{"id":"doi:10.48550/arxiv.2605.24869","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2605.24869","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.2605.24869","any_repository_has_fulltext":true},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5101237514","display_name":"Yunao Zheng","orcid":"https://orcid.org/0009-0002-4583-3607"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Zheng, Yunao","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5120722033","display_name":"Guoyang Xia","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Xia, Guoyang","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5137065372","display_name":"Xiaojie Wang","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Wang, Xiaojie","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"last","author":{"id":"https://openalex.org/A5137066925","display_name":"Lei Ren","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Ren, Lei","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/T11714","display_name":"Multimodal Machine Learning Applications","score":0.4569000005722046,"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"}},"topics":[{"id":"https://openalex.org/T11714","display_name":"Multimodal Machine Learning Applications","score":0.4569000005722046,"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"}},{"id":"https://openalex.org/T10028","display_name":"Topic Modeling","score":0.22300000488758087,"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":0.14300000667572021,"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/perplexity","display_name":"Perplexity","score":0.7674999833106995},{"id":"https://openalex.org/keywords/inference","display_name":"Inference","score":0.555400013923645},{"id":"https://openalex.org/keywords/hash-function","display_name":"Hash function","score":0.4140999913215637},{"id":"https://openalex.org/keywords/language-model","display_name":"Language model","score":0.4099999964237213},{"id":"https://openalex.org/keywords/transformer","display_name":"Transformer","score":0.40310001373291016},{"id":"https://openalex.org/keywords/source-code","display_name":"Source code","score":0.40209999680519104},{"id":"https://openalex.org/keywords/question-answering","display_name":"Question answering","score":0.3790000081062317},{"id":"https://openalex.org/keywords/encoding","display_name":"Encoding (memory)","score":0.37860000133514404}],"concepts":[{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.7824000120162964},{"id":"https://openalex.org/C100279451","wikidata":"https://www.wikidata.org/wiki/Q372193","display_name":"Perplexity","level":3,"score":0.7674999833106995},{"id":"https://openalex.org/C2776214188","wikidata":"https://www.wikidata.org/wiki/Q408386","display_name":"Inference","level":2,"score":0.555400013923645},{"id":"https://openalex.org/C204321447","wikidata":"https://www.wikidata.org/wiki/Q30642","display_name":"Natural language processing","level":1,"score":0.5156000256538391},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.4754999876022339},{"id":"https://openalex.org/C99138194","wikidata":"https://www.wikidata.org/wiki/Q183427","display_name":"Hash function","level":2,"score":0.4140999913215637},{"id":"https://openalex.org/C137293760","wikidata":"https://www.wikidata.org/wiki/Q3621696","display_name":"Language model","level":2,"score":0.4099999964237213},{"id":"https://openalex.org/C66322947","wikidata":"https://www.wikidata.org/wiki/Q11658","display_name":"Transformer","level":3,"score":0.40310001373291016},{"id":"https://openalex.org/C43126263","wikidata":"https://www.wikidata.org/wiki/Q128751","display_name":"Source code","level":2,"score":0.40209999680519104},{"id":"https://openalex.org/C44291984","wikidata":"https://www.wikidata.org/wiki/Q1074173","display_name":"Question answering","level":2,"score":0.3790000081062317},{"id":"https://openalex.org/C125411270","wikidata":"https://www.wikidata.org/wiki/Q18653","display_name":"Encoding (memory)","level":2,"score":0.37860000133514404},{"id":"https://openalex.org/C2778112365","wikidata":"https://www.wikidata.org/wiki/Q3511065","display_name":"Sequence (biology)","level":2,"score":0.37059998512268066},{"id":"https://openalex.org/C80444323","wikidata":"https://www.wikidata.org/wiki/Q2878974","display_name":"Theoretical computer science","level":1,"score":0.3702999949455261},{"id":"https://openalex.org/C2777210771","wikidata":"https://www.wikidata.org/wiki/Q4927124","display_name":"Block (permutation group theory)","level":2,"score":0.33239999413490295},{"id":"https://openalex.org/C28093023","wikidata":"https://www.wikidata.org/wiki/Q175168","display_name":"Engram","level":2,"score":0.33079999685287476},{"id":"https://openalex.org/C36503486","wikidata":"https://www.wikidata.org/wiki/Q11235244","display_name":"Domain (mathematical analysis)","level":2,"score":0.314300000667572},{"id":"https://openalex.org/C100660578","wikidata":"https://www.wikidata.org/wiki/Q18733","display_name":"Recall","level":2,"score":0.31130000948905945},{"id":"https://openalex.org/C2776760102","wikidata":"https://www.wikidata.org/wiki/Q5139990","display_name":"Code (set theory)","level":3,"score":0.31029999256134033},{"id":"https://openalex.org/C206134035","wikidata":"https://www.wikidata.org/wiki/Q811525","display_name":"Treebank","level":3,"score":0.29820001125335693},{"id":"https://openalex.org/C118505674","wikidata":"https://www.wikidata.org/wiki/Q42586063","display_name":"Encoder","level":2,"score":0.2930000126361847},{"id":"https://openalex.org/C119857082","wikidata":"https://www.wikidata.org/wiki/Q2539","display_name":"Machine learning","level":1,"score":0.2824000120162964},{"id":"https://openalex.org/C114466953","wikidata":"https://www.wikidata.org/wiki/Q6034165","display_name":"Initialization","level":2,"score":0.2750999927520752},{"id":"https://openalex.org/C80156102","wikidata":"https://www.wikidata.org/wiki/Q788036","display_name":"Resolver","level":3,"score":0.26840001344680786},{"id":"https://openalex.org/C11413529","wikidata":"https://www.wikidata.org/wiki/Q8366","display_name":"Algorithm","level":1,"score":0.2540999948978424}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.48550/arxiv.2605.24869","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2605.24869","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.2605.24869","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2605.24869","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":[{"display_name":"Quality Education","id":"https://metadata.un.org/sdg/4","score":0.5580026507377625}],"awards":[],"funders":[],"has_content":{"pdf":false,"grobid_xml":false},"content_urls":null,"referenced_works_count":0,"referenced_works":[],"related_works":[],"abstract_inverted_index":{"Sequence":[0],"modeling":[1],"requires":[2],"both":[3,15],"compositional":[4],"reasoning":[5],"and":[6,35,54,69,82,92,117,126,143],"local":[7],"static":[8],"knowledge":[9,96],"retrieval,":[10],"yet":[11],"standard":[12],"Transformers":[13],"handle":[14],"through":[16],"dense":[17],"computation.":[18],"Engram":[19,83],"partially":[20],"decouples":[21],"retrieval":[22],"from":[23,51],"the":[24,64,107],"backbone,":[25],"but":[26],"its":[27],"token-based":[28],"keys":[29],"remain":[30],"tied":[31],"to":[32,72,101,134],"text":[33],"tokenization":[34],"hash":[36],"compression.":[37],"We":[38],"propose":[39],"Lngram,":[40],"a":[41],"latent-space":[42],"conditional":[43],"memory":[44,144],"module":[45],"that":[46,129],"learns":[47],"discrete":[48],"symbols":[49],"directly":[50],"hidden":[52],"states":[53],"performs":[55],"N-gram":[56],"lookup":[57],"over":[58],"these":[59],"symbols.":[60],"This":[61],"design":[62],"removes":[63],"dependence":[65],"on":[66,115],"tokenizer":[67],"IDs":[68],"naturally":[70],"extends":[71],"non-text":[73],"modalities.":[74],"In":[75],"our":[76],"evaluated":[77],"settings,":[78],"Lngram":[79,130],"outperforms":[80],"Transformer":[81],"baselines,":[84],"consistently":[85],"reduces":[86],"perplexity":[87],"in":[88],"long-context":[89],"language":[90],"modeling,":[91],"effectively":[93],"injects":[94],"domain":[95],"when":[97],"added":[98],"post":[99],"hoc":[100],"pretrained":[102],"models.":[103],"Joint":[104],"training":[105],"with":[106,124,140],"backbone":[108],"further":[109],"surpasses":[110],"full":[111],"fine-tuning,":[112],"while":[113],"experiments":[114],"vision-language":[116],"vision-language-action":[118],"tasks":[119],"show":[120],"overall":[121],"gains.":[122],"Analyses":[123],"LogitLens":[125],"CKA":[127],"suggest":[128],"enables":[131],"prediction-relevant":[132],"information":[133],"emerge":[135],"earlier,":[136],"increasing":[137],"effective":[138],"depth":[139],"limited":[141],"inference":[142],"overhead.":[145],"Code":[146],"is":[147],"available":[148],"at":[149],"https://github.com/zyaaa-ux/Lngram.":[150]},"counts_by_year":[],"updated_date":"2026-07-01T06:00:48.157686","created_date":"2026-05-27T00:00:00"}