{"id":"https://openalex.org/W7164018556","doi":"https://doi.org/10.1109/access.2026.3701952","title":"Reinforcement Learning With Verifier Guidance and Penalty Shaping for Vietnamese Summarization Using Small Language Models","display_name":"Reinforcement Learning With Verifier Guidance and Penalty Shaping for Vietnamese Summarization Using Small Language Models","publication_year":2026,"publication_date":"2026-01-01","ids":{"openalex":"https://openalex.org/W7164018556","doi":"https://doi.org/10.1109/access.2026.3701952"},"language":"en","primary_location":{"id":"doi:10.1109/access.2026.3701952","is_oa":true,"landing_page_url":"https://doi.org/10.1109/access.2026.3701952","pdf_url":null,"source":{"id":"https://openalex.org/S2485537415","display_name":"IEEE Access","issn_l":"2169-3536","issn":["2169-3536"],"is_oa":true,"is_in_doaj":true,"is_core":true,"host_organization":"https://openalex.org/P4310319808","host_organization_name":"Institute of Electrical and Electronics Engineers","host_organization_lineage":["https://openalex.org/P4310319808"],"host_organization_lineage_names":["Institute of Electrical and Electronics Engineers"],"type":"journal"},"license":"cc-by","license_id":"https://openalex.org/licenses/cc-by","version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"IEEE Access","raw_type":"journal-article"},"type":"article","indexed_in":["crossref","doaj"],"open_access":{"is_oa":true,"oa_status":"gold","oa_url":"https://doi.org/10.1109/access.2026.3701952","any_repository_has_fulltext":false},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5113119386","display_name":"Thi Thu Phuong Tran","orcid":null},"institutions":[{"id":"https://openalex.org/I64519617","display_name":"Hanoi University","ror":"https://ror.org/01mxx0e62","country_code":"VN","type":"education","lineage":["https://openalex.org/I64519617"]}],"countries":["VN"],"is_corresponding":false,"raw_author_name":"Thi Thu Phuong Tran","raw_affiliation_strings":["Hanoi Metropolitan University, Nghia Do, Hanoi, Vietnam"],"raw_orcid":"https://orcid.org/0009-0001-9501-4514","affiliations":[{"raw_affiliation_string":"Hanoi Metropolitan University, Nghia Do, Hanoi, Vietnam","institution_ids":["https://openalex.org/I64519617"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5011813480","display_name":"Tr\u1ecbnh Ho\u00e0ng V\u0169","orcid":null},"institutions":[{"id":"https://openalex.org/I177233841","display_name":"Vietnam National University, Hanoi","ror":"https://ror.org/02jmfj006","country_code":"VN","type":"education","lineage":["https://openalex.org/I177233841"]}],"countries":["VN"],"is_corresponding":false,"raw_author_name":"Trinh Hoang Vu","raw_affiliation_strings":["VNU University of Engineering and Technology, Vietnam National University, Cau Giay, Hanoi, Vietnam"],"raw_orcid":"https://orcid.org/0009-0000-3936-2250","affiliations":[{"raw_affiliation_string":"VNU University of Engineering and Technology, Vietnam National University, Cau Giay, Hanoi, Vietnam","institution_ids":["https://openalex.org/I177233841"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5138242674","display_name":"Van Vinh Nguyen","orcid":null},"institutions":[{"id":"https://openalex.org/I177233841","display_name":"Vietnam National University, Hanoi","ror":"https://ror.org/02jmfj006","country_code":"VN","type":"education","lineage":["https://openalex.org/I177233841"]}],"countries":["VN"],"is_corresponding":false,"raw_author_name":"Van Vinh Nguyen","raw_affiliation_strings":["VNU University of Engineering and Technology, Vietnam National University, Cau Giay, Hanoi, Vietnam"],"raw_orcid":"https://orcid.org/0009-0004-0912-6328","affiliations":[{"raw_affiliation_string":"VNU University of Engineering and Technology, Vietnam National University, Cau Giay, Hanoi, Vietnam","institution_ids":["https://openalex.org/I177233841"]}]},{"author_position":"last","author":{"id":"https://openalex.org/A5103974499","display_name":"Phuong Thai Nguyen","orcid":null},"institutions":[{"id":"https://openalex.org/I177233841","display_name":"Vietnam National University, Hanoi","ror":"https://ror.org/02jmfj006","country_code":"VN","type":"education","lineage":["https://openalex.org/I177233841"]}],"countries":["VN"],"is_corresponding":false,"raw_author_name":"Phuong Thai Nguyen","raw_affiliation_strings":["VNU University of Engineering and Technology, Vietnam National University, Cau Giay, Hanoi, Vietnam"],"raw_orcid":"https://orcid.org/0009-0006-2033-7788","affiliations":[{"raw_affiliation_string":"VNU University of Engineering and Technology, Vietnam National University, Cau Giay, Hanoi, Vietnam","institution_ids":["https://openalex.org/I177233841"]}]}],"institutions":[],"countries_distinct_count":1,"institutions_distinct_count":2,"corresponding_author_ids":[],"corresponding_institution_ids":[],"apc_list":{"value":1850,"currency":"USD","value_usd":1850},"apc_paid":{"value":1850,"currency":"USD","value_usd":1850},"fwci":0.0,"has_fulltext":false,"cited_by_count":0,"citation_normalized_percentile":{"value":0.73222448,"is_in_top_1_percent":false,"is_in_top_10_percent":false},"cited_by_percentile_year":null,"biblio":{"volume":"14","issue":null,"first_page":"89301","last_page":"89322"},"is_retracted":false,"is_paratext":false,"is_xpac":false,"primary_topic":{"id":"https://openalex.org/T10028","display_name":"Topic Modeling","score":0.7037000060081482,"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":0.7037000060081482,"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":0.0746999979019165,"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/T10201","display_name":"Speech Recognition and Synthesis","score":0.04569999873638153,"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/automatic-summarization","display_name":"Automatic summarization","score":0.8406000137329102},{"id":"https://openalex.org/keywords/vietnamese","display_name":"Vietnamese","score":0.7027999758720398},{"id":"https://openalex.org/keywords/reinforcement-learning","display_name":"Reinforcement learning","score":0.6814000010490417},{"id":"https://openalex.org/keywords/language-model","display_name":"Language model","score":0.35510000586509705},{"id":"https://openalex.org/keywords/language-acquisition","display_name":"Language acquisition","score":0.2727999985218048}],"concepts":[{"id":"https://openalex.org/C170858558","wikidata":"https://www.wikidata.org/wiki/Q1394144","display_name":"Automatic summarization","level":2,"score":0.8406000137329102},{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.8259999752044678},{"id":"https://openalex.org/C103621254","wikidata":"https://www.wikidata.org/wiki/Q9199","display_name":"Vietnamese","level":2,"score":0.7027999758720398},{"id":"https://openalex.org/C97541855","wikidata":"https://www.wikidata.org/wiki/Q830687","display_name":"Reinforcement learning","level":2,"score":0.6814000010490417},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.5723000168800354},{"id":"https://openalex.org/C204321447","wikidata":"https://www.wikidata.org/wiki/Q30642","display_name":"Natural language processing","level":1,"score":0.4230000078678131},{"id":"https://openalex.org/C137293760","wikidata":"https://www.wikidata.org/wiki/Q3621696","display_name":"Language model","level":2,"score":0.35510000586509705},{"id":"https://openalex.org/C119857082","wikidata":"https://www.wikidata.org/wiki/Q2539","display_name":"Machine learning","level":1,"score":0.321399986743927},{"id":"https://openalex.org/C74672266","wikidata":"https://www.wikidata.org/wiki/Q815859","display_name":"Language acquisition","level":2,"score":0.2727999985218048},{"id":"https://openalex.org/C67186912","wikidata":"https://www.wikidata.org/wiki/Q367664","display_name":"Data modeling","level":2,"score":0.2614000141620636},{"id":"https://openalex.org/C175154964","wikidata":"https://www.wikidata.org/wiki/Q380077","display_name":"Task analysis","level":3,"score":0.25920000672340393},{"id":"https://openalex.org/C183322885","wikidata":"https://www.wikidata.org/wiki/Q17007702","display_name":"Context model","level":3,"score":0.25040000677108765}],"mesh":[],"locations_count":2,"locations":[{"id":"doi:10.1109/access.2026.3701952","is_oa":true,"landing_page_url":"https://doi.org/10.1109/access.2026.3701952","pdf_url":null,"source":{"id":"https://openalex.org/S2485537415","display_name":"IEEE Access","issn_l":"2169-3536","issn":["2169-3536"],"is_oa":true,"is_in_doaj":true,"is_core":true,"host_organization":"https://openalex.org/P4310319808","host_organization_name":"Institute of Electrical and Electronics Engineers","host_organization_lineage":["https://openalex.org/P4310319808"],"host_organization_lineage_names":["Institute of Electrical and Electronics Engineers"],"type":"journal"},"license":"cc-by","license_id":"https://openalex.org/licenses/cc-by","version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"IEEE Access","raw_type":"journal-article"},{"id":"pmh:oai:doaj.org/article:4550289a08474a36a06879d168bfd644","is_oa":false,"landing_page_url":"https://doaj.org/article/4550289a08474a36a06879d168bfd644","pdf_url":null,"source":{"id":"https://openalex.org/S4306401280","display_name":"DOAJ (DOAJ: Directory of Open Access Journals)","issn_l":null,"issn":null,"is_oa":false,"is_in_doaj":false,"is_core":false,"host_organization":null,"host_organization_name":null,"host_organization_lineage":[],"host_organization_lineage_names":[],"type":"repository"},"license":null,"license_id":null,"version":"submittedVersion","is_accepted":false,"is_published":false,"raw_source_name":"IEEE Access, Vol 14, Pp 89301-89322 (2026)","raw_type":"article"}],"best_oa_location":{"id":"doi:10.1109/access.2026.3701952","is_oa":true,"landing_page_url":"https://doi.org/10.1109/access.2026.3701952","pdf_url":null,"source":{"id":"https://openalex.org/S2485537415","display_name":"IEEE Access","issn_l":"2169-3536","issn":["2169-3536"],"is_oa":true,"is_in_doaj":true,"is_core":true,"host_organization":"https://openalex.org/P4310319808","host_organization_name":"Institute of Electrical and Electronics Engineers","host_organization_lineage":["https://openalex.org/P4310319808"],"host_organization_lineage_names":["Institute of Electrical and Electronics Engineers"],"type":"journal"},"license":"cc-by","license_id":"https://openalex.org/licenses/cc-by","version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"IEEE Access","raw_type":"journal-article"},"sustainable_development_goals":[],"awards":[{"id":"https://openalex.org/G673697367","display_name":null,"funder_award_id":"QG.23.73","funder_id":"https://openalex.org/F4320315892","funder_display_name":"University of Science, Vietnam National University, Hanoi"}],"funders":[{"id":"https://openalex.org/F4320315892","display_name":"University of Science, Vietnam National University, Hanoi","ror":"https://ror.org/05w54hk79"}],"has_content":{"grobid_xml":false,"pdf":false},"content_urls":null,"referenced_works_count":0,"referenced_works":[],"related_works":[],"abstract_inverted_index":{"Reinforcement":[0],"learning":[1],"(RL)":[2],"can":[3,14],"optimize":[4],"abstractive":[5],"summarization":[6,203],"toward":[7],"task":[8],"quality":[9,34],"criteria,":[10],"but":[11],"imperfect":[12],"rewards":[13],"cause":[15],"reward":[16],"hacking,":[17],"factual":[18],"drift,":[19],"and":[20,35,58,72,78,112,124,133,140,159,177],"unstable":[21],"updates.":[22],"This":[23],"issue":[24],"is":[25],"acute":[26],"for":[27,75,109,116,155],"Vietnamese":[28,145,202],"summarization,":[29],"where":[30],"references":[31],"of":[32,185],"high":[33],"human":[36,178],"preference":[37],"data":[38],"at":[39],"scale":[40],"are":[41],"costly,":[42],"while":[43,163,205],"deployment":[44],"often":[45],"favors":[46],"small":[47,201],"language":[48],"models.":[49],"We":[50,87],"propose":[51],"an":[52],"RL":[53],"framework":[54,62,199],"with":[55,98,101,152],"verifier":[56],"guidance":[57],"penalty":[59],"shaping.":[60],"The":[61],"uses":[63],"the":[64,95,183,194,198],"frozen":[65],"MultiEvalSumViet2":[66],"evaluator":[67,96],"to":[68],"score":[69],"source":[70],"document":[71],"summary":[73],"pairs":[74],"faithfulness,":[76],"coherence,":[77],"relevance,":[79],"then":[80],"converts":[81],"these":[82],"scores":[83],"into":[84],"bounded":[85],"rewards.":[86],"compare":[88],"Verified":[89,99],"Reward":[90,100],"(VR),":[91],"which":[92,104],"directly":[93],"optimizes":[94],"reward,":[97],"Penalty":[102],"(VRP),":[103],"adds":[105],"a":[106],"hard":[107],"gate":[108],"clear":[110],"failures":[111],"calibrated":[113],"soft":[114],"penalties":[115],"borderline":[117],"outputs.We":[118],"evaluate":[119],"Proximal":[120],"Policy":[121,127],"Optimization":[122,128],"(PPO)":[123],"Group":[125],"Relative":[126],"(GRPO)":[129],"on":[130,208],"ViT5-large,":[131],"Llama-3.2-1B,":[132],"Qwen3-0.6B":[134],"under":[135,161],"supervised":[136],"fine":[137],"tuning":[138],"(SFT)":[139],"without":[141],"SFT":[142,158],"(noSFT)":[143],"initialization.":[144],"news":[146,195],"experiments":[147],"show":[148],"improved":[149],"rubric":[150],"quality,":[151],"strong":[153],"gains":[154],"ViT5":[156],"after":[157],"Qwen3":[160],"noSFT,":[162],"Llama":[164],"benefits":[165],"more":[166],"clearly":[167],"from":[168],"GRPO":[169],"than":[170],"PPO.":[171],"Component":[172],"ablations,":[173],"95%":[174],"confidence":[175],"intervals,":[176],"review":[179],"reliability":[180],"analysis":[181],"clarify":[182],"contribution":[184],"VRP.":[186],"A":[187],"625-example":[188],"WikiLingua":[189],"test":[190],"suggests":[191],"transfer":[192],"beyond":[193],"domain.":[196],"Overall,":[197],"improves":[200],"models":[204],"reducing":[206],"reliance":[207],"reference":[209],"summaries.":[210]},"counts_by_year":[],"updated_date":"2026-06-20T20:08:15.867695","created_date":"2026-06-10T00:00:00"}