{"id":"https://openalex.org/W7139959733","doi":"https://doi.org/10.48550/arxiv.2603.18074","title":"Lightweight Adaptation for LLM-based Technical Service Agent: Latent Logic Augmentation and Robust Noise Reduction","display_name":"Lightweight Adaptation for LLM-based Technical Service Agent: Latent Logic Augmentation and Robust Noise Reduction","publication_year":2026,"publication_date":"2026-03-18","ids":{"openalex":"https://openalex.org/W7139959733","doi":"https://doi.org/10.48550/arxiv.2603.18074"},"language":null,"primary_location":{"id":"doi:10.48550/arxiv.2603.18074","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2603.18074","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":"cc-by","license_id":"https://openalex.org/licenses/cc-by","version":null,"is_accepted":false,"is_published":false,"raw_source_name":null,"raw_type":"article"},"type":"preprint","indexed_in":["datacite"],"open_access":{"is_oa":true,"oa_status":"green","oa_url":"https://doi.org/10.48550/arxiv.2603.18074","any_repository_has_fulltext":true},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5130224086","display_name":"Yi Yu","orcid":null},"institutions":[],"countries":[],"is_corresponding":true,"raw_author_name":"Yu, Yi","raw_affiliation_strings":[],"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5130215761","display_name":"Junzhuo Ma","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Ma, Junzhuo","raw_affiliation_strings":[],"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5130251958","display_name":"Chenghuang Shen","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Shen, Chenghuang","raw_affiliation_strings":[],"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5050215381","display_name":"Xingyan Liu","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Liu, Xingyan","raw_affiliation_strings":[],"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5130235128","display_name":"Jing Gu","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Gu, Jing","raw_affiliation_strings":[],"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5130249200","display_name":"Hangyi Sun","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Sun, Hangyi","raw_affiliation_strings":[],"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5130232709","display_name":"Guangquan Hu","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Hu, Guangquan","raw_affiliation_strings":[],"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5130234379","display_name":"Jianfeng Liu","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Liu, Jianfeng","raw_affiliation_strings":[],"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5124130589","display_name":"Weiting Liu","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Liu, Weiting","raw_affiliation_strings":[],"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5130232714","display_name":"Mingyue Pu","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Pu, Mingyue","raw_affiliation_strings":[],"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5130228308","display_name":"Yu Wang","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Wang, Yu","raw_affiliation_strings":[],"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5015218544","display_name":"Zhengdong Xiao","orcid":"https://orcid.org/0000-0001-7536-1052"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Xiao, Zhengdong","raw_affiliation_strings":[],"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5130230933","display_name":"Rui Xie","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Xie, Rui","raw_affiliation_strings":[],"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5130241998","display_name":"Longjiu Luo","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Luo, Longjiu","raw_affiliation_strings":[],"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5100730844","display_name":"Yu Wang","orcid":"https://orcid.org/0000-0001-7894-6523"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Wang, Qianrong","raw_affiliation_strings":[],"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5017561045","display_name":"G. Cui","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Cui, Gurong","raw_affiliation_strings":[],"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5130233520","display_name":"Honglin Qiao","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Qiao, Honglin","raw_affiliation_strings":[],"affiliations":[]},{"author_position":"last","author":{"id":"https://openalex.org/A5130215177","display_name":"Wenlian Lu","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Lu, Wenlian","raw_affiliation_strings":[],"affiliations":[]}],"institutions":[],"countries_distinct_count":0,"institutions_distinct_count":18,"corresponding_author_ids":["https://openalex.org/A5130224086"],"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":0.29339998960494995,"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":0.29339998960494995,"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/T12026","display_name":"Explainable Artificial Intelligence (XAI)","score":0.08479999750852585,"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/T11636","display_name":"Artificial Intelligence in Healthcare and Education","score":0.08259999752044678,"subfield":{"id":"https://openalex.org/subfields/2718","display_name":"Health Informatics"},"field":{"id":"https://openalex.org/fields/27","display_name":"Medicine"},"domain":{"id":"https://openalex.org/domains/4","display_name":"Health Sciences"}}],"keywords":[{"id":"https://openalex.org/keywords/ambiguity","display_name":"Ambiguity","score":0.597599983215332},{"id":"https://openalex.org/keywords/adaptation","display_name":"Adaptation (eye)","score":0.5368000268936157},{"id":"https://openalex.org/keywords/noise","display_name":"Noise (video)","score":0.49729999899864197},{"id":"https://openalex.org/keywords/key","display_name":"Key (lock)","score":0.482699990272522},{"id":"https://openalex.org/keywords/stability","display_name":"Stability (learning theory)","score":0.46320000290870667},{"id":"https://openalex.org/keywords/bridge","display_name":"Bridge (graph theory)","score":0.427700012922287},{"id":"https://openalex.org/keywords/reinforcement-learning","display_name":"Reinforcement learning","score":0.41920000314712524},{"id":"https://openalex.org/keywords/construct","display_name":"Construct (python library)","score":0.39410001039505005}],"concepts":[{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.7602999806404114},{"id":"https://openalex.org/C2780522230","wikidata":"https://www.wikidata.org/wiki/Q1140419","display_name":"Ambiguity","level":2,"score":0.597599983215332},{"id":"https://openalex.org/C139807058","wikidata":"https://www.wikidata.org/wiki/Q352374","display_name":"Adaptation (eye)","level":2,"score":0.5368000268936157},{"id":"https://openalex.org/C119857082","wikidata":"https://www.wikidata.org/wiki/Q2539","display_name":"Machine learning","level":1,"score":0.5236999988555908},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.5055999755859375},{"id":"https://openalex.org/C99498987","wikidata":"https://www.wikidata.org/wiki/Q2210247","display_name":"Noise (video)","level":3,"score":0.49729999899864197},{"id":"https://openalex.org/C26517878","wikidata":"https://www.wikidata.org/wiki/Q228039","display_name":"Key (lock)","level":2,"score":0.482699990272522},{"id":"https://openalex.org/C112972136","wikidata":"https://www.wikidata.org/wiki/Q7595718","display_name":"Stability (learning theory)","level":2,"score":0.46320000290870667},{"id":"https://openalex.org/C100776233","wikidata":"https://www.wikidata.org/wiki/Q2532492","display_name":"Bridge (graph theory)","level":2,"score":0.427700012922287},{"id":"https://openalex.org/C97541855","wikidata":"https://www.wikidata.org/wiki/Q830687","display_name":"Reinforcement learning","level":2,"score":0.41920000314712524},{"id":"https://openalex.org/C2780801425","wikidata":"https://www.wikidata.org/wiki/Q5164392","display_name":"Construct (python library)","level":2,"score":0.39410001039505005},{"id":"https://openalex.org/C2780378061","wikidata":"https://www.wikidata.org/wiki/Q25351891","display_name":"Service (business)","level":2,"score":0.3840000033378601},{"id":"https://openalex.org/C206345919","wikidata":"https://www.wikidata.org/wiki/Q20380951","display_name":"Resource (disambiguation)","level":2,"score":0.3174999952316284},{"id":"https://openalex.org/C106189395","wikidata":"https://www.wikidata.org/wiki/Q176789","display_name":"Markov decision process","level":3,"score":0.31299999356269836},{"id":"https://openalex.org/C63479239","wikidata":"https://www.wikidata.org/wiki/Q7353546","display_name":"Robustness (evolution)","level":3,"score":0.2897999882698059},{"id":"https://openalex.org/C124101348","wikidata":"https://www.wikidata.org/wiki/Q172491","display_name":"Data mining","level":1,"score":0.27469998598098755},{"id":"https://openalex.org/C13662910","wikidata":"https://www.wikidata.org/wiki/Q193139","display_name":"Trajectory","level":2,"score":0.272599995136261},{"id":"https://openalex.org/C65965080","wikidata":"https://www.wikidata.org/wiki/Q1806885","display_name":"Latent variable model","level":3,"score":0.27239999175071716},{"id":"https://openalex.org/C163294075","wikidata":"https://www.wikidata.org/wiki/Q581861","display_name":"Noise reduction","level":2,"score":0.26179999113082886},{"id":"https://openalex.org/C111335779","wikidata":"https://www.wikidata.org/wiki/Q3454686","display_name":"Reduction (mathematics)","level":2,"score":0.2596000134944916}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.48550/arxiv.2603.18074","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2603.18074","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":"cc-by","license_id":"https://openalex.org/licenses/cc-by","version":null,"is_accepted":false,"is_published":null,"raw_source_name":null,"raw_type":"article"}],"best_oa_location":{"id":"doi:10.48550/arxiv.2603.18074","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2603.18074","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":"cc-by","license_id":"https://openalex.org/licenses/cc-by","version":null,"is_accepted":false,"is_published":false,"raw_source_name":null,"raw_type":"article"},"sustainable_development_goals":[],"awards":[],"funders":[],"has_content":{"pdf":false,"grobid_xml":false},"content_urls":null,"referenced_works_count":0,"referenced_works":[],"related_works":[],"abstract_inverted_index":{"Adapting":[0],"Large":[1],"Language":[2],"Models":[3],"in":[4,18],"complex":[5],"technical":[6,228],"service":[7,182,229],"domains":[8],"is":[9,48],"constrained":[10],"by":[11,51,133],"the":[12,22,27,52,96,108,131,139,168,223],"absence":[13],"of":[14,29,110],"explicit":[15],"cognitive":[16],"chains":[17],"human":[19],"demonstrations":[20],"and":[21,42,55,67,90,101,195,202],"inherent":[23],"ambiguity":[24],"arising":[25],"from":[26,37],"diversity":[28],"valid":[30],"responses.":[31],"These":[32,105],"limitations":[33],"severely":[34],"hinder":[35],"agents":[36],"internalizing":[38],"latent":[39,102],"decision":[40,103],"dynamics":[41],"generalizing":[43],"effectively.":[44],"Moreover,":[45],"practical":[46,224],"adaptation":[47,75],"often":[49],"impeded":[50],"prohibitive":[53],"resource":[54],"time":[56],"costs":[57],"associated":[58],"with":[59,156,218],"standard":[60,173,215],"training":[61,220],"paradigms.":[62],"To":[63],"overcome":[64],"these":[65],"challenges":[66],"guarantee":[68],"computational":[69,169],"efficiency,":[70],"we":[71],"propose":[72],"a":[73,120,126,147,157],"lightweight":[74,158],"framework":[76,192],"comprising":[77],"three":[78],"key":[79],"contributions.":[80],"(1)":[81],"Latent":[82,199],"Logic":[83,200],"Augmentation:":[84],"We":[85,118,145],"introduce":[86],"Planning-Aware":[87],"Trajectory":[88],"Modeling":[89],"Decision":[91],"Reasoning":[92],"Augmentation":[93,201],"to":[94,129,161,172,214],"bridge":[95],"gap":[97],"between":[98],"surface-level":[99],"supervision":[100],"logic.":[104],"approaches":[106],"strengthen":[107],"stability":[109,194],"Supervised":[111],"Fine-Tuning":[112],"alignment.":[113],"(2)":[114],"Robust":[115,203],"Noise":[116,204],"Reduction:":[117],"construct":[119],"Multiple":[121],"Ground":[122],"Truths":[123],"dataset":[124],"through":[125,198],"dual-filtering":[127],"method":[128],"reduce":[130],"noise":[132],"validating":[134],"diverse":[135,187],"responses,":[136],"thereby":[137],"capturing":[138],"semantic":[140],"diversity.":[141],"(3)":[142],"Lightweight":[143],"Adaptation:":[144],"design":[146],"Hybrid":[148,208],"Reward":[149,209],"mechanism":[150,210],"that":[151,190],"fuses":[152],"an":[153],"LLM-based":[154],"judge":[155],"relevance-based":[159],"Reranker":[160],"distill":[162],"high-fidelity":[163],"reward":[164],"signals":[165],"while":[166],"reducing":[167],"cost":[170],"compared":[171],"LLM-as-a-Judge":[174],"reinforcement":[175],"learning.":[176],"Empirical":[177],"evaluations":[178],"on":[179],"real-world":[180],"Cloud":[181],"tasks,":[183],"conducted":[184],"across":[185],"semantically":[186],"settings,":[188],"demonstrate":[189],"our":[191,207],"achieves":[193,211],"performance":[196],"gains":[197],"Reduction.":[205],"Concurrently,":[206],"alignment":[212],"comparable":[213],"LLM-as-a-judge":[216],"methods":[217],"reduced":[219],"time,":[221],"underscoring":[222],"value":[225],"for":[226],"deploying":[227],"agents.":[230]},"counts_by_year":[],"updated_date":"2026-03-21T06:36:02.116451","created_date":"2026-03-21T00:00:00"}