{"id":"https://openalex.org/W7125767870","doi":"https://doi.org/10.48550/arxiv.2601.18129","title":"Typhoon-S: Minimal Open Post-Training for Sovereign Large Language Models","display_name":"Typhoon-S: Minimal Open Post-Training for Sovereign Large Language Models","publication_year":2026,"publication_date":"2026-01-26","ids":{"openalex":"https://openalex.org/W7125767870","doi":"https://doi.org/10.48550/arxiv.2601.18129"},"language":null,"primary_location":{"id":"doi:10.48550/arxiv.2601.18129","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2601.18129","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,"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"},"type":"preprint","indexed_in":["datacite"],"open_access":{"is_oa":true,"oa_status":"green","oa_url":"https://doi.org/10.48550/arxiv.2601.18129","any_repository_has_fulltext":true},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5124008894","display_name":"Kunat Pipatanakul","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Pipatanakul, Kunat","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"last","author":{"id":"https://openalex.org/A5088392357","display_name":"Pittawat Taveekitworachai","orcid":"https://orcid.org/0000-0002-6824-2634"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Taveekitworachai, Pittawat","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/T10028","display_name":"Topic Modeling","score":0.18960000574588776,"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/T10028","display_name":"Topic Modeling","score":0.18960000574588776,"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":0.08889999985694885,"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/T13910","display_name":"Computational and Text Analysis Methods","score":0.08299999684095383,"subfield":{"id":"https://openalex.org/subfields/3300","display_name":"General Social Sciences"},"field":{"id":"https://openalex.org/fields/33","display_name":"Social Sciences"},"domain":{"id":"https://openalex.org/domains/2","display_name":"Social Sciences"}}],"keywords":[{"id":"https://openalex.org/keywords/transparency","display_name":"Transparency (behavior)","score":0.5437999963760376},{"id":"https://openalex.org/keywords/adversarial-system","display_name":"Adversarial system","score":0.41589999198913574},{"id":"https://openalex.org/keywords/reinforcement-learning","display_name":"Reinforcement learning","score":0.4083000123500824},{"id":"https://openalex.org/keywords/pipeline","display_name":"Pipeline (software)","score":0.38029998540878296},{"id":"https://openalex.org/keywords/language-model","display_name":"Language model","score":0.3707999885082245},{"id":"https://openalex.org/keywords/gatekeeping","display_name":"Gatekeeping","score":0.3617999851703644},{"id":"https://openalex.org/keywords/scale","display_name":"Scale (ratio)","score":0.337799996137619},{"id":"https://openalex.org/keywords/knowledge-base","display_name":"Knowledge base","score":0.3319000005722046}],"concepts":[{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.6955999732017517},{"id":"https://openalex.org/C2780233690","wikidata":"https://www.wikidata.org/wiki/Q535347","display_name":"Transparency (behavior)","level":2,"score":0.5437999963760376},{"id":"https://openalex.org/C37736160","wikidata":"https://www.wikidata.org/wiki/Q1801315","display_name":"Adversarial system","level":2,"score":0.41589999198913574},{"id":"https://openalex.org/C97541855","wikidata":"https://www.wikidata.org/wiki/Q830687","display_name":"Reinforcement learning","level":2,"score":0.4083000123500824},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.3849000036716461},{"id":"https://openalex.org/C43521106","wikidata":"https://www.wikidata.org/wiki/Q2165493","display_name":"Pipeline (software)","level":2,"score":0.38029998540878296},{"id":"https://openalex.org/C137293760","wikidata":"https://www.wikidata.org/wiki/Q3621696","display_name":"Language model","level":2,"score":0.3707999885082245},{"id":"https://openalex.org/C73380752","wikidata":"https://www.wikidata.org/wiki/Q609087","display_name":"Gatekeeping","level":2,"score":0.3617999851703644},{"id":"https://openalex.org/C2778755073","wikidata":"https://www.wikidata.org/wiki/Q10858537","display_name":"Scale (ratio)","level":2,"score":0.337799996137619},{"id":"https://openalex.org/C4554734","wikidata":"https://www.wikidata.org/wiki/Q593744","display_name":"Knowledge base","level":2,"score":0.3319000005722046},{"id":"https://openalex.org/C36503486","wikidata":"https://www.wikidata.org/wiki/Q11235244","display_name":"Domain (mathematical analysis)","level":2,"score":0.31150001287460327},{"id":"https://openalex.org/C2778137410","wikidata":"https://www.wikidata.org/wiki/Q2732820","display_name":"Government (linguistics)","level":2,"score":0.2922999858856201},{"id":"https://openalex.org/C187691185","wikidata":"https://www.wikidata.org/wiki/Q2020720","display_name":"Grid","level":2,"score":0.28700000047683716},{"id":"https://openalex.org/C2777063073","wikidata":"https://www.wikidata.org/wiki/Q1553237","display_name":"Settlement (finance)","level":3,"score":0.2858000099658966},{"id":"https://openalex.org/C186644900","wikidata":"https://www.wikidata.org/wiki/Q194152","display_name":"Parsing","level":2,"score":0.28220000863075256},{"id":"https://openalex.org/C199776023","wikidata":"https://www.wikidata.org/wiki/Q202875","display_name":"Negotiation","level":2,"score":0.27000001072883606},{"id":"https://openalex.org/C119857082","wikidata":"https://www.wikidata.org/wiki/Q2539","display_name":"Machine learning","level":1,"score":0.2637999951839447},{"id":"https://openalex.org/C2164484","wikidata":"https://www.wikidata.org/wiki/Q5170150","display_name":"Core (optical fiber)","level":2,"score":0.2619999945163727},{"id":"https://openalex.org/C2775924081","wikidata":"https://www.wikidata.org/wiki/Q55608371","display_name":"Control (management)","level":2,"score":0.25699999928474426},{"id":"https://openalex.org/C204321447","wikidata":"https://www.wikidata.org/wiki/Q30642","display_name":"Natural language processing","level":1,"score":0.2540999948978424},{"id":"https://openalex.org/C2780522230","wikidata":"https://www.wikidata.org/wiki/Q1140419","display_name":"Ambiguity","level":2,"score":0.2524000108242035},{"id":"https://openalex.org/C94922259","wikidata":"https://www.wikidata.org/wiki/Q33215","display_name":"Constructed language","level":2,"score":0.25130000710487366}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.48550/arxiv.2601.18129","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2601.18129","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,"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":null,"raw_source_name":null,"raw_type":"Preprint"}],"best_oa_location":{"id":"doi:10.48550/arxiv.2601.18129","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2601.18129","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,"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"},"sustainable_development_goals":[{"score":0.40829959511756897,"id":"https://metadata.un.org/sdg/4","display_name":"Quality Education"}],"awards":[],"funders":[],"has_content":{"grobid_xml":false,"pdf":false},"content_urls":null,"referenced_works_count":0,"referenced_works":[],"related_works":[],"abstract_inverted_index":{"Large":[0],"language":[1],"models":[2,10,184,187],"(LLMs)":[3],"have":[4],"progressed":[5],"rapidly;":[6],"however,":[7],"most":[8],"state-of-the-art":[9],"are":[11,25],"trained":[12],"and":[13,22,24,39,63,70,77,102,119,142,153,163,181,220,244],"evaluated":[14],"primarily":[15],"in":[16,50,116],"high-resource":[17],"languages":[18,118],"such":[19],"as":[20,168],"English":[21],"Chinese,":[23],"often":[26],"developed":[27],"by":[28],"a":[29,44,52,95,99,151,169,211,231,247],"small":[30],"number":[31],"of":[32,65,203,241],"organizations":[33],"with":[34,188,198,210],"access":[35],"to":[36,93,108],"large-scale":[37,143],"compute":[38],"data.":[40],"This":[41],"gatekeeping":[42],"creates":[43],"practical":[45,248],"barrier":[46],"for":[47],"sovereign":[48,104,252],"settings":[49],"which":[51],"regional-":[53],"or":[54,57,135],"national-scale":[55],"institution":[56],"domain":[58],"owner":[59],"must":[60],"retain":[61],"control":[62],"understanding":[64],"model":[66,97],"weights,":[67],"training":[68],"data,":[69],"deployment":[71],"while":[72,223],"operating":[73],"under":[74,254],"limited":[75],"resources":[76],"strict":[78],"transparency":[79],"constraints.":[80],"To":[81],"this":[82],"end,":[83],"we":[84,173],"identify":[85],"two":[86],"core":[87],"requirements:":[88],"(1)":[89],"adoptability,":[90],"the":[91,106,207,238],"ability":[92,107],"transform":[94],"base":[96,183],"into":[98,185],"general-purpose":[100,182],"assistant,":[101],"(2)":[103],"capability,":[105],"perform":[109],"high-stakes,":[110],"region-specific":[111],"tasks":[112],"(e.g.,":[113],"legal":[114,218],"reasoning":[115,219],"local":[117],"cultural":[120],"knowledge).":[121],"We":[122,147,192],"investigate":[123],"whether":[124],"these":[125],"requirements":[126],"can":[127,236],"be":[128],"achieved":[129],"without":[130],"scaling":[131],"massive":[132],"instruction":[133,242],"corpora":[134],"relying":[136],"on":[137],"complex":[138],"preference":[139],"tuning":[140],"pipelines":[141],"reinforcement":[144],"fine-tuning":[145],"(RFT).":[146],"present":[148],"Typhoon":[149],"S,":[150],"minimal":[152],"open":[154],"post-training":[155,234],"recipe":[156],"that":[157,175,195,205,230],"combines":[158],"supervised":[159],"fine-tuning,":[160],"on-policy":[161],"distillation,":[162],"small-scale":[164,196],"RFT.":[165],"Using":[166],"Thai":[167,217],"representative":[170],"case":[171],"study,":[172],"demonstrate":[174],"our":[176],"approach":[177],"transforms":[178],"both":[179],"sovereign-adapted":[180],"instruction-tuned":[186],"strong":[189],"general":[190,225],"performance.":[191],"further":[193],"show":[194],"RFT":[197],"InK-GRPO":[199],"--":[200,215],"an":[201],"extension":[202],"GRPO":[204,208],"augments":[206],"loss":[209,214],"next-word":[212],"prediction":[213],"improves":[216],"Thai-specific":[221],"knowledge":[222],"preserving":[224],"capabilities.":[226],"Our":[227],"results":[228],"suggest":[229],"carefully":[232],"designed":[233],"strategy":[235],"reduce":[237],"required":[239],"scale":[240],"data":[243],"computation,":[245],"providing":[246],"path":[249],"toward":[250],"high-quality":[251],"LLMs":[253],"academic-scale":[255],"resources.":[256]},"counts_by_year":[],"updated_date":"2026-07-28T07:46:37.118299","created_date":"2026-01-28T00:00:00"}