{"id":"https://openalex.org/W7165678886","doi":"https://doi.org/10.48550/arxiv.2606.21262","title":"ARCO: Adaptive Rubric with Co-Evolution for Multi-Step LLM-Based Agents","display_name":"ARCO: Adaptive Rubric with Co-Evolution for Multi-Step LLM-Based Agents","publication_year":2026,"publication_date":"2026-06-19","ids":{"openalex":"https://openalex.org/W7165678886","doi":"https://doi.org/10.48550/arxiv.2606.21262"},"language":null,"primary_location":{"id":"doi:10.48550/arxiv.2606.21262","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2606.21262","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":"Preprint"},"type":"preprint","indexed_in":["datacite"],"open_access":{"is_oa":true,"oa_status":"green","oa_url":"https://doi.org/10.48550/arxiv.2606.21262","any_repository_has_fulltext":true},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5108891390","display_name":"Zihang Tian","orcid":"https://orcid.org/0009-0009-8292-5384"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Tian, Zihang","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5101658564","display_name":"Jingsen Zhang","orcid":"https://orcid.org/0000-0003-2997-3386"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Zhang, Jingsen","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5052597908","display_name":"R K Li","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Li, Rui","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5139159412","display_name":"Xiaohe Bo","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Bo, Xiaohe","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5139166877","display_name":"Yuanzi Li","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Li, Yuanzi","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"last","author":{"id":"https://openalex.org/A5139214315","display_name":"Xu Chen","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Chen, Xu","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/T10462","display_name":"Reinforcement Learning in Robotics","score":0.4041999876499176,"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":0.4041999876499176,"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":0.2996000051498413,"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/T11714","display_name":"Multimodal Machine Learning Applications","score":0.03060000017285347,"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"}}],"keywords":[{"id":"https://openalex.org/keywords/rubric","display_name":"Rubric","score":0.9083999991416931},{"id":"https://openalex.org/keywords/interpretability","display_name":"Interpretability","score":0.8184000253677368},{"id":"https://openalex.org/keywords/trajectory","display_name":"Trajectory","score":0.4339999854564667},{"id":"https://openalex.org/keywords/parameterized-complexity","display_name":"Parameterized complexity","score":0.4189000129699707},{"id":"https://openalex.org/keywords/function","display_name":"Function (biology)","score":0.4099999964237213},{"id":"https://openalex.org/keywords/reinforcement-learning","display_name":"Reinforcement learning","score":0.4050999879837036}],"concepts":[{"id":"https://openalex.org/C111640148","wikidata":"https://www.wikidata.org/wiki/Q847349","display_name":"Rubric","level":2,"score":0.9083999991416931},{"id":"https://openalex.org/C2781067378","wikidata":"https://www.wikidata.org/wiki/Q17027399","display_name":"Interpretability","level":2,"score":0.8184000253677368},{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.6424999833106995},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.43779999017715454},{"id":"https://openalex.org/C13662910","wikidata":"https://www.wikidata.org/wiki/Q193139","display_name":"Trajectory","level":2,"score":0.4339999854564667},{"id":"https://openalex.org/C119857082","wikidata":"https://www.wikidata.org/wiki/Q2539","display_name":"Machine learning","level":1,"score":0.42149999737739563},{"id":"https://openalex.org/C165464430","wikidata":"https://www.wikidata.org/wiki/Q1570441","display_name":"Parameterized complexity","level":2,"score":0.4189000129699707},{"id":"https://openalex.org/C14036430","wikidata":"https://www.wikidata.org/wiki/Q3736076","display_name":"Function (biology)","level":2,"score":0.4099999964237213},{"id":"https://openalex.org/C97541855","wikidata":"https://www.wikidata.org/wiki/Q830687","display_name":"Reinforcement learning","level":2,"score":0.4050999879837036},{"id":"https://openalex.org/C2776036281","wikidata":"https://www.wikidata.org/wiki/Q48769818","display_name":"Constraint (computer-aided design)","level":2,"score":0.37540000677108765},{"id":"https://openalex.org/C112313634","wikidata":"https://www.wikidata.org/wiki/Q7886648","display_name":"Complement (music)","level":5,"score":0.29120001196861267},{"id":"https://openalex.org/C136197465","wikidata":"https://www.wikidata.org/wiki/Q1729295","display_name":"Variety (cybernetics)","level":2,"score":0.266400009393692},{"id":"https://openalex.org/C124681953","wikidata":"https://www.wikidata.org/wiki/Q339062","display_name":"Decomposition","level":2,"score":0.26089999079704285},{"id":"https://openalex.org/C124101348","wikidata":"https://www.wikidata.org/wiki/Q172491","display_name":"Data mining","level":1,"score":0.2563999891281128}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.48550/arxiv.2606.21262","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2606.21262","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":"Preprint"}],"best_oa_location":{"id":"doi:10.48550/arxiv.2606.21262","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2606.21262","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":"Preprint"},"sustainable_development_goals":[{"score":0.4995824098587036,"display_name":"Peace, Justice and strong institutions","id":"https://metadata.un.org/sdg/16"}],"awards":[],"funders":[],"has_content":{"pdf":false,"grobid_xml":false},"content_urls":null,"referenced_works_count":0,"referenced_works":[],"related_works":[],"abstract_inverted_index":{"Reinforcement":[0],"learning":[1],"for":[2,178],"multi-step":[3],"LLM":[4],"agents":[5],"often":[6],"relies":[7],"on":[8,123],"scalar":[9],"rewards":[10,25,103],"that":[11,81,89,127,167],"indicate":[12],"success":[13],"but":[14,31],"cannot":[15],"explain":[16],"why":[17],"a":[18,44,63,68,73,78,86],"trajectory":[19,37,95],"is":[20],"good":[21],"or":[22],"bad.":[23],"Rubric-based":[24],"improve":[26],"interpretability":[27],"through":[28],"natural-language":[29],"criteria,":[30,84],"existing":[32],"methods":[33],"score":[34,87],"at":[35,136,187],"the":[36,41,53,99,105,117,128,132,137,151],"level":[38],"and":[39,52,85,116,131,143,161,164,176,183],"freeze":[40],"scorer":[42],"behind":[43],"closed-source":[45],"judge,":[46],"leaving":[47],"step-level":[48,92,112],"credit":[49,109],"assignment":[50,110],"unresolved":[51],"judge":[54],"itself":[55],"static.":[56],"We":[57],"propose":[58],"ARCO":[59,149],"(Adaptive":[60],"Rubric":[61],"CO-evolution),":[62],"rubric":[64,129],"framework":[65],"in":[66,154],"which":[67],"same-scale":[69],"model":[70],"$\u03bc$":[71,115],"shares":[72],"backbone":[74],"with":[75,145],"two":[76,146],"heads:":[77],"generation":[79],"head":[80,88],"produces":[82],"per-step":[83],"predicts":[90],"rubric-conditioned":[91],"rewards.":[93],"A":[94],"decomposition":[96],"constraint":[97],"ties":[98],"sum":[100],"of":[101],"step":[102],"to":[104,173],"terminal":[106],"outcome,":[107],"enabling":[108],"without":[111],"labels,":[113],"while":[114],"policy":[118],"$\u03c0$":[119],"are":[120,170,185],"jointly":[121],"updated":[122],"on-policy":[124],"data":[125,184],"so":[126],"content":[130],"scoring":[133],"function":[134],"co-evolve":[135],"parameter":[138],"level.":[139],"Across":[140],"HotpotQA,":[141],"2WikiMultiHopQA,":[142],"MuSiQue":[144],"open-source":[147],"backbones,":[148],"improves":[150],"best":[152],"EM":[153],"every":[155],"setting":[156],"over":[157],"strong":[158],"outcome-,":[159],"rubric-,":[160],"process-reward":[162],"baselines,":[163],"analyses":[165],"show":[166],"its":[168],"rubrics":[169],"step-specific,":[171],"robust":[172],"design":[174],"choices,":[175],"useful":[177],"diagnosing":[179],"agent":[180],"behavior.":[181],"Codes":[182],"available":[186],"https://github.com/zihangtian/ARCO.":[188]},"counts_by_year":[],"updated_date":"2026-07-01T06:00:48.157686","created_date":"2026-06-24T00:00:00"}