{"id":"https://openalex.org/W7204179234","doi":"https://doi.org/10.18653/2026.emnlp-main.268","title":"Agent-G2: Gaussian Guidance for Agentic Reinforcement Learning","display_name":"Agent-G2: Gaussian Guidance for Agentic Reinforcement Learning","publication_year":2026,"publication_date":"2026-01-01","ids":{"openalex":"https://openalex.org/W7204179234","doi":"https://doi.org/10.18653/2026.emnlp-main.268"},"language":"en","primary_location":{"id":"doi:10.18653/2026.emnlp-main.268","doi":"https://doi.org/10.18653/2026.emnlp-main.268","is_oa":true,"landing_page_url":"https://doi.org/10.18653/2026.emnlp-main.268","pdf_url":"https://aclanthology.org/2026.emnlp-main.268.pdf","source":{"id":"https://openalex.org/S4363608991","display_name":"Conference on Empirical Methods in Natural Language Processing (EMNLP)","issn_l":null,"issn":null,"is_oa":false,"is_in_doaj":false,"is_core":false,"listed_in":[],"host_organization":null,"host_organization_name":null,"host_organization_lineage":[],"host_organization_lineage_names":[],"type":"conference"},"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"Proceedings of the 2026 Conference on Empirical Methods in Natural Language Processing","raw_type":"proceedings-article"},"volume":null,"type":"conference-paper","indexed_in":["arxiv","crossref","datacite"],"open_access":{"is_oa":true,"oa_status":"bronze","oa_url":"https://aclanthology.org/2026.emnlp-main.268.pdf","any_repository_has_fulltext":true},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5148066048","display_name":"Zixuan Wang","orcid":null,"observed_orcids":[]},"institutions":[{"id":"https://openalex.org/I76130692","display_name":"Zhejiang University","ror":"https://ror.org/00a2xv884","country_code":"CN","type":"education","lineage":["https://openalex.org/I76130692"]},{"id":"https://openalex.org/I98301712","display_name":"Baidu (China)","ror":"https://ror.org/03vs3wt56","country_code":"CN","type":"company","lineage":["https://openalex.org/I98301712"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Zixuan Wang","raw_affiliation_strings":["Baidu Inc","Zhejiang University"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Baidu Inc","institution_ids":["https://openalex.org/I98301712"]},{"raw_affiliation_string":"Zhejiang University","institution_ids":["https://openalex.org/I76130692"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5030548874","display_name":"Yanrui Miao","orcid":null,"observed_orcids":[]},"institutions":[{"id":"https://openalex.org/I76130692","display_name":"Zhejiang University","ror":"https://ror.org/00a2xv884","country_code":"CN","type":"education","lineage":["https://openalex.org/I76130692"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Yanrui Miao","raw_affiliation_strings":["Zhejiang University"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Zhejiang University","institution_ids":["https://openalex.org/I76130692"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5146013130","display_name":"Zhengxi Lu","orcid":null,"observed_orcids":[]},"institutions":[{"id":"https://openalex.org/I76130692","display_name":"Zhejiang University","ror":"https://ror.org/00a2xv884","country_code":"CN","type":"education","lineage":["https://openalex.org/I76130692"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Zhengxi Lu","raw_affiliation_strings":["Zhejiang University"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Zhejiang University","institution_ids":["https://openalex.org/I76130692"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5130059365","display_name":"Teng Pan","orcid":null,"observed_orcids":[]},"institutions":[{"id":"https://openalex.org/I76130692","display_name":"Zhejiang University","ror":"https://ror.org/00a2xv884","country_code":"CN","type":"education","lineage":["https://openalex.org/I76130692"]},{"id":"https://openalex.org/I98301712","display_name":"Baidu (China)","ror":"https://ror.org/03vs3wt56","country_code":"CN","type":"company","lineage":["https://openalex.org/I98301712"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Teng Pan","raw_affiliation_strings":["Baidu Inc","Zhejiang University"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Baidu Inc","institution_ids":["https://openalex.org/I98301712"]},{"raw_affiliation_string":"Zhejiang University","institution_ids":["https://openalex.org/I76130692"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5110784690","display_name":"Yiwen Qiu","orcid":null,"observed_orcids":[]},"institutions":[{"id":"https://openalex.org/I76130692","display_name":"Zhejiang University","ror":"https://ror.org/00a2xv884","country_code":"CN","type":"education","lineage":["https://openalex.org/I76130692"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Yiwen Qiu","raw_affiliation_strings":["Zhejiang University"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Zhejiang University","institution_ids":["https://openalex.org/I76130692"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5135652744","display_name":"Hongxing Li","orcid":null,"observed_orcids":[]},"institutions":[{"id":"https://openalex.org/I76130692","display_name":"Zhejiang University","ror":"https://ror.org/00a2xv884","country_code":"CN","type":"education","lineage":["https://openalex.org/I76130692"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Hongxing Li","raw_affiliation_strings":["Zhejiang University"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Zhejiang University","institution_ids":["https://openalex.org/I76130692"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5125162301","display_name":"Peng Qiu","orcid":null,"observed_orcids":[]},"institutions":[{"id":"https://openalex.org/I98301712","display_name":"Baidu (China)","ror":"https://ror.org/03vs3wt56","country_code":"CN","type":"company","lineage":["https://openalex.org/I98301712"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Peng Qiu","raw_affiliation_strings":["Baidu Inc"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Baidu Inc","institution_ids":["https://openalex.org/I98301712"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5126177900","display_name":"Ruiqing Zhang","orcid":null,"observed_orcids":[]},"institutions":[{"id":"https://openalex.org/I98301712","display_name":"Baidu (China)","ror":"https://ror.org/03vs3wt56","country_code":"CN","type":"company","lineage":["https://openalex.org/I98301712"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Ruiqing Zhang","raw_affiliation_strings":["Baidu Inc"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Baidu Inc","institution_ids":["https://openalex.org/I98301712"]}]},{"author_position":"last","author":{"id":"https://openalex.org/A5148129839","display_name":"Yongliang Shen","orcid":null,"observed_orcids":[]},"institutions":[{"id":"https://openalex.org/I76130692","display_name":"Zhejiang University","ror":"https://ror.org/00a2xv884","country_code":"CN","type":"education","lineage":["https://openalex.org/I76130692"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Yongliang Shen","raw_affiliation_strings":["Zhejiang University"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Zhejiang University","institution_ids":["https://openalex.org/I76130692"]}]}],"institutions":[],"countries_distinct_count":1,"institutions_distinct_count":2,"corresponding_author_ids":[],"corresponding_institution_ids":[],"apc_list":null,"apc_paid":null,"fwci":0.0,"has_fulltext":true,"cited_by_count":0,"citation_normalized_percentile":{"value":0.67678736,"is_in_top_1_percent":false,"is_in_top_10_percent":false},"cited_by_percentile_year":null,"biblio":{"volume":null,"issue":null,"first_page":"6096","last_page":"6111"},"is_retracted":false,"is_paratext":false,"is_xpac":false,"primary_topic":{"id":"https://openalex.org/T10462","display_name":"Reinforcement Learning in Robotics","score":0.9984999895095825,"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/T10462","display_name":"Reinforcement Learning in Robotics","score":0.9984999895095825,"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/T10906","display_name":"AI-based Problem Solving and Planning","score":0.00019999999494757503,"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/T11307","display_name":"Domain Adaptation and Few-Shot Learning","score":9.999999747378752e-05,"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/agentic-reinforcement-learning","display_name":"agentic reinforcement learning","score":0.8970000147819519},{"id":"https://openalex.org/keywords/reward-sparsity","display_name":"reward sparsity","score":0.8669999837875366},{"id":"https://openalex.org/keywords/long-horizon-tasks","display_name":"long-horizon tasks","score":0.9089999794960022},{"id":"https://openalex.org/keywords/online-estimation","display_name":"online estimation","score":0.42899999022483826},{"id":"https://openalex.org/keywords/alfworld","display_name":"ALFWorld","score":0.8140000104904175},{"id":"https://openalex.org/keywords/webshop","display_name":"WebShop","score":0.8119999766349792},{"id":"https://openalex.org/keywords/llm-based-agents","display_name":"LLM agents","score":0.7139999866485596}],"concepts":[{"id":"https://openalex.org/C13662910","wikidata":"https://www.wikidata.org/wiki/Q193139","display_name":"Trajectory","level":2,"score":0.7156000137329102},{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.6941999793052673},{"id":"https://openalex.org/C97541855","wikidata":"https://www.wikidata.org/wiki/Q830687","display_name":"Reinforcement learning","level":2,"score":0.6721000075340271},{"id":"https://openalex.org/C163716315","wikidata":"https://www.wikidata.org/wiki/Q901177","display_name":"Gaussian","level":2,"score":0.5914000272750854},{"id":"https://openalex.org/C61326573","wikidata":"https://www.wikidata.org/wiki/Q1496376","display_name":"Gaussian process","level":3,"score":0.5103999972343445},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.4742000102996826},{"id":"https://openalex.org/C2780451532","wikidata":"https://www.wikidata.org/wiki/Q759676","display_name":"Task (project management)","level":2,"score":0.43799999356269836},{"id":"https://openalex.org/C119857082","wikidata":"https://www.wikidata.org/wiki/Q2539","display_name":"Machine learning","level":1,"score":0.4117000102996826},{"id":"https://openalex.org/C12725497","wikidata":"https://www.wikidata.org/wiki/Q810247","display_name":"Baseline (sea)","level":2,"score":0.39559999108314514},{"id":"https://openalex.org/C48103436","wikidata":"https://www.wikidata.org/wiki/Q599031","display_name":"State (computer science)","level":2,"score":0.35740000009536743},{"id":"https://openalex.org/C11413529","wikidata":"https://www.wikidata.org/wiki/Q8366","display_name":"Algorithm","level":1,"score":0.3346000015735626},{"id":"https://openalex.org/C126255220","wikidata":"https://www.wikidata.org/wiki/Q141495","display_name":"Mathematical optimization","level":1,"score":0.3321000039577484},{"id":"https://openalex.org/C106189395","wikidata":"https://www.wikidata.org/wiki/Q176789","display_name":"Markov decision process","level":3,"score":0.2816999852657318},{"id":"https://openalex.org/C192209626","wikidata":"https://www.wikidata.org/wiki/Q190909","display_name":"Focus (optics)","level":2,"score":0.2678000032901764},{"id":"https://openalex.org/C2779436431","wikidata":"https://www.wikidata.org/wiki/Q30672407","display_name":"Policy learning","level":2,"score":0.2662999927997589},{"id":"https://openalex.org/C112972136","wikidata":"https://www.wikidata.org/wiki/Q7595718","display_name":"Stability (learning theory)","level":2,"score":0.25920000672340393},{"id":"https://openalex.org/C2779463800","wikidata":"https://www.wikidata.org/wiki/Q5062222","display_name":"Center (category theory)","level":2,"score":0.25839999318122864},{"id":"https://openalex.org/C2779795794","wikidata":"https://www.wikidata.org/wiki/Q7315343","display_name":"Reset (finance)","level":2,"score":0.2549999952316284},{"id":"https://openalex.org/C28719098","wikidata":"https://www.wikidata.org/wiki/Q44946","display_name":"Point (geometry)","level":2,"score":0.25270000100135803}],"mesh":[],"locations_count":3,"locations":[{"id":"doi:10.18653/2026.emnlp-main.268","doi":"https://doi.org/10.18653/2026.emnlp-main.268","is_oa":true,"landing_page_url":"https://doi.org/10.18653/2026.emnlp-main.268","pdf_url":"https://aclanthology.org/2026.emnlp-main.268.pdf","source":{"id":"https://openalex.org/S4363608991","display_name":"Conference on Empirical Methods in Natural Language Processing (EMNLP)","issn_l":null,"issn":null,"is_oa":false,"is_in_doaj":false,"is_core":false,"listed_in":[],"host_organization":null,"host_organization_name":null,"host_organization_lineage":[],"host_organization_lineage_names":[],"type":"conference"},"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"Proceedings of the 2026 Conference on Empirical Methods in Natural Language Processing","raw_type":"proceedings-article"},{"id":"pmh:oai:arXiv.org:2608.23318","doi":null,"is_oa":true,"landing_page_url":"https://arxiv.org/abs/2608.23318","pdf_url":"https://arxiv.org/pdf/2608.23318","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,"listed_in":[],"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":"submittedVersion","is_accepted":false,"is_published":false,"raw_source_name":null,"raw_type":"text"},{"id":"doi:10.48550/arxiv.2608.23318","doi":"https://doi.org/10.48550/arxiv.2608.23318","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2608.23318","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,"listed_in":[],"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"}],"best_oa_location":{"id":"doi:10.18653/2026.emnlp-main.268","doi":"https://doi.org/10.18653/2026.emnlp-main.268","is_oa":true,"landing_page_url":"https://doi.org/10.18653/2026.emnlp-main.268","pdf_url":"https://aclanthology.org/2026.emnlp-main.268.pdf","source":{"id":"https://openalex.org/S4363608991","display_name":"Conference on Empirical Methods in Natural Language Processing (EMNLP)","issn_l":null,"issn":null,"is_oa":false,"is_in_doaj":false,"is_core":false,"listed_in":[],"host_organization":null,"host_organization_name":null,"host_organization_lineage":[],"host_organization_lineage_names":[],"type":"conference"},"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"Proceedings of the 2026 Conference on Empirical Methods in Natural Language Processing","raw_type":"proceedings-article"},"sustainable_development_goals":[],"sustainable_development_goals_aurora":[{"id":"https://metadata.un.org/sdg/16","score":0.5426289439201355,"display_name":"Peace, justice, and strong institutions"}],"study_designs":[],"awards":[],"funders":[{"id":"https://openalex.org/F4320318547","display_name":"Baidu","ror":"https://ror.org/03vs3wt56"},{"id":"https://openalex.org/F4320321001","display_name":"National Natural Science Foundation of China","ror":"https://ror.org/01h0zpd94"},{"id":"https://openalex.org/F4320329860","display_name":"National Science and Technology Major Project","ror":null}],"has_content":{"pdf":true,"grobid_xml":true},"content_urls":{"pdf":"https://content.openalex.org/works/W7204179234.pdf","grobid_xml":"https://content.openalex.org/works/W7204179234.grobid-xml"},"referenced_works_count":0,"referenced_works":[],"related_works":["https://openalex.org/W7202134906","https://openalex.org/W7204181592","https://openalex.org/W7208147649","https://openalex.org/W2159666783","https://openalex.org/W4362683553","https://openalex.org/W7160991704","https://openalex.org/W7160662643","https://openalex.org/W7148414856","https://openalex.org/W3096551261","https://openalex.org/W7214951477"],"abstract_inverted_index":{"Hint-based":[0],"reinforcement":[1],"learning":[2],"addresses":[3],"reward":[4],"sparsity":[5],"in":[6],"long-horizon":[7],"agentic":[8],"tasks":[9],"by":[10,174],"retaining":[11],"a":[12,26,49,78,96,104,115,141],"prefix":[13],"of":[14,39,70,80,187],"an":[15],"expert":[16],"trajectory":[17,41],"before":[18],"each":[19],"rollout,":[20],"letting":[21],"the":[22,34,40,68,89,110,148,165,184],"policy":[23,129],"explore":[24],"from":[25,114,124],"state":[27],"closer":[28],"to":[29,42],"success.Its":[30],"effectiveness":[31],"hinges":[32],"on":[33,156,160,172],"guidance":[35,76,106],"depth:":[36],"how":[37],"much":[38],"keep.Existing":[43],"methods":[44],"treat":[45],"this":[46],"depth":[47,111,137],"as":[48],"deterministic":[50],"scalar.Scheduled":[51],"approaches":[52],"share":[53],"one":[54],"value":[55],"across":[56],"samples":[57],"and":[58,119,147,158,169],"ignore":[59],"per-task":[60],"heterogeneity;":[61],"persample":[62],"probing":[63],"estimates":[64],"it":[65],"separately":[66],"at":[67,95,181,192],"cost":[69,186],"extra":[71],"rollouts.We":[72],"find":[73],"that":[74,108],"useful":[75],"occupies":[77],"band":[79,90],"depths":[81],"whose":[82,117],"informativeness":[83],"profile":[84],"is":[85,190],"approximately":[86],"Gaussian":[87,105,116],"around":[88],"center,":[91],"rather":[92],"than":[93],"concentrating":[94],"single":[97],"optimal":[98],"point.We":[99],"propose":[100],"Agent-G":[101,154],"2":[102,155,163],",":[103],"framework":[107],"draws":[109],"per":[112],"task":[113],"center":[118,139],"spread":[120,149],"are":[121],"estimated":[122],"online":[123],"rollouts":[125,134],"already":[126],"collected":[127],"for":[128],"optimization,":[130],"requiring":[131],"no":[132],"probe":[133],"or":[135],"learned":[136],"predictor.The":[138],"combines":[140],"global":[142],"baseline":[143],"with":[144],"per-cluster":[145],"difficulty,":[146],"tracks":[150],"within-cluster":[151],"variance.We":[152],"evaluate":[153],"ALFWorld":[157,173],"WebShop":[159],"Qwen2.5-1.5B/":[161],"7B-Instruct.Agent-G":[162],"outperforms":[164],"strongest":[166],"hint-based,":[167],"hint-free,":[168],"Aux-RL":[170],"baselines":[171],"2.3":[175],"/":[176,178],"3.9":[177],"7.4":[179],"points":[180],"under":[182],"one-third":[183],"rollout":[185],"per-sample":[188],"probing.Code":[189],"available":[191],"https://github.com/ZJU-REAL/Agent-G2.":[193]},"counts_by_year":[],"updated_date":"2026-10-31T12:09:11.096373","created_date":"2026-08-26T00:00:00"}