{"id":"https://openalex.org/W4415089340","doi":"https://doi.org/10.48550/arxiv.2509.11963","title":"ToolRM: Outcome Reward Models for Tool-Calling Large Language Models","display_name":"ToolRM: Outcome Reward Models for Tool-Calling Large Language Models","publication_year":2025,"publication_date":"2025-09-15","ids":{"openalex":"https://openalex.org/W4415089340","doi":"https://doi.org/10.48550/arxiv.2509.11963"},"language":"en","primary_location":{"id":"pmh:oai:arXiv.org:2509.11963","doi":null,"is_oa":true,"landing_page_url":"http://arxiv.org/abs/2509.11963","pdf_url":"https://arxiv.org/pdf/2509.11963","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":null,"license_id":null,"version":"submittedVersion","is_accepted":false,"is_published":false,"raw_source_name":null,"raw_type":"text"},"volume":null,"type":"preprint","indexed_in":["arxiv","datacite"],"open_access":{"is_oa":true,"oa_status":"green","oa_url":"https://arxiv.org/pdf/2509.11963","any_repository_has_fulltext":true},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5076160843","display_name":"Mayank Agarwal","orcid":"https://orcid.org/0000-0002-6374-6737","observed_orcids":["https://orcid.org/0000-0002-6374-6737","https://orcid.org/0009-0004-0582-2836"]},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Agarwal, Mayank","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5031546123","display_name":"Ibrahim Abdelaziz","orcid":"https://orcid.org/0000-0003-1449-5115","observed_orcids":["https://orcid.org/0000-0003-1449-5115"]},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Abdelaziz, Ibrahim","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5078166693","display_name":"Kinjal Basu","orcid":"https://orcid.org/0000-0002-4091-0119","observed_orcids":["https://orcid.org/0000-0002-4091-0119"]},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Basu, Kinjal","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5013192617","display_name":"Merve Unuvar","orcid":null,"observed_orcids":[]},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Unuvar, Merve","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5023268703","display_name":"Luis Lastras","orcid":"https://orcid.org/0009-0006-2792-220X","observed_orcids":["https://orcid.org/0009-0006-2792-220X"]},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Lastras, Luis A.","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5058002550","display_name":"Yara Rizk","orcid":"https://orcid.org/0000-0001-7475-2385","observed_orcids":["https://orcid.org/0000-0001-7475-2385"]},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Rizk, Yara","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"last","author":{"id":"https://openalex.org/A5003720552","display_name":"Pavan Kapanipathi","orcid":"https://orcid.org/0000-0003-0494-3279","observed_orcids":["https://orcid.org/0000-0003-0494-3279"]},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Kapanipathi, Pavan","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/T10181","display_name":"Natural Language Processing Techniques","score":0.326200008392334,"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/T10181","display_name":"Natural Language Processing Techniques","score":0.326200008392334,"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/T10028","display_name":"Topic Modeling","score":0.28929999470710754,"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/T10462","display_name":"Reinforcement Learning in Robotics","score":0.0828000009059906,"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/tool-calling","display_name":"tool calling","score":0.7059999704360962},{"id":"https://openalex.org/keywords/reward-modeling","display_name":"reward modeling","score":0.7039999961853027},{"id":"https://openalex.org/keywords/large-language-models","display_name":"large language models","score":0.8510000109672546},{"id":"https://openalex.org/keywords/benchmark","display_name":"benchmark","score":0.4259999990463257},{"id":"https://openalex.org/keywords/synthetic-training-data","display_name":"synthetic training data","score":0.7400000095367432},{"id":"https://openalex.org/keywords/open-weight-llms","display_name":"open-weight LLMs","score":0.8619999885559082},{"id":"https://openalex.org/keywords/best-of-n-sampling","display_name":"best-of-N sampling","score":0.8100000023841858},{"id":"https://openalex.org/keywords/input-noise-robustness","display_name":"input noise robustness","score":0.7570000290870667}],"concepts":[{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.7454000115394592},{"id":"https://openalex.org/C137293760","wikidata":"https://www.wikidata.org/wiki/Q3621696","display_name":"Language model","level":2,"score":0.6309000253677368},{"id":"https://openalex.org/C79581498","wikidata":"https://www.wikidata.org/wiki/Q1367530","display_name":"Suite","level":2,"score":0.5914999842643738},{"id":"https://openalex.org/C63479239","wikidata":"https://www.wikidata.org/wiki/Q7353546","display_name":"Robustness (evolution)","level":3,"score":0.571399986743927},{"id":"https://openalex.org/C148220186","wikidata":"https://www.wikidata.org/wiki/Q7111912","display_name":"Outcome (game theory)","level":2,"score":0.560699999332428},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.5568000078201294},{"id":"https://openalex.org/C119857082","wikidata":"https://www.wikidata.org/wiki/Q2539","display_name":"Machine learning","level":1,"score":0.5515999794006348},{"id":"https://openalex.org/C26517878","wikidata":"https://www.wikidata.org/wiki/Q228039","display_name":"Key (lock)","level":2,"score":0.4196999967098236},{"id":"https://openalex.org/C185798385","wikidata":"https://www.wikidata.org/wiki/Q1161707","display_name":"Benchmark (surveying)","level":2,"score":0.3970000147819519},{"id":"https://openalex.org/C195324797","wikidata":"https://www.wikidata.org/wiki/Q33742","display_name":"Natural language","level":2,"score":0.387800008058548},{"id":"https://openalex.org/C2779439875","wikidata":"https://www.wikidata.org/wiki/Q1078276","display_name":"Natural language understanding","level":3,"score":0.3156999945640564},{"id":"https://openalex.org/C67186912","wikidata":"https://www.wikidata.org/wiki/Q367664","display_name":"Data modeling","level":2,"score":0.2939999997615814},{"id":"https://openalex.org/C179603123","wikidata":"https://www.wikidata.org/wiki/Q1941921","display_name":"Modeling language","level":3,"score":0.28380000591278076},{"id":"https://openalex.org/C204321447","wikidata":"https://www.wikidata.org/wiki/Q30642","display_name":"Natural language processing","level":1,"score":0.2827000021934509},{"id":"https://openalex.org/C143661069","wikidata":"https://www.wikidata.org/wiki/Q670713","display_name":"Reward system","level":2,"score":0.28200000524520874},{"id":"https://openalex.org/C66024118","wikidata":"https://www.wikidata.org/wiki/Q1122506","display_name":"Computational model","level":2,"score":0.274399995803833},{"id":"https://openalex.org/C115051666","wikidata":"https://www.wikidata.org/wiki/Q6522493","display_name":"Ranging","level":2,"score":0.26080000400543213}],"mesh":[],"locations_count":2,"locations":[{"id":"pmh:oai:arXiv.org:2509.11963","doi":null,"is_oa":true,"landing_page_url":"http://arxiv.org/abs/2509.11963","pdf_url":"https://arxiv.org/pdf/2509.11963","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":null,"license_id":null,"version":"submittedVersion","is_accepted":false,"is_published":false,"raw_source_name":null,"raw_type":"text"},{"id":"doi:10.48550/arxiv.2509.11963","doi":"https://doi.org/10.48550/arxiv.2509.11963","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2509.11963","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":null,"license_id":null,"version":null,"is_accepted":false,"is_published":null,"raw_source_name":null,"raw_type":"Preprint"}],"best_oa_location":{"id":"pmh:oai:arXiv.org:2509.11963","doi":null,"is_oa":true,"landing_page_url":"http://arxiv.org/abs/2509.11963","pdf_url":"https://arxiv.org/pdf/2509.11963","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":null,"license_id":null,"version":"submittedVersion","is_accepted":false,"is_published":false,"raw_source_name":null,"raw_type":"text"},"sustainable_development_goals":[],"sustainable_development_goals_aurora":[],"study_designs":[],"awards":[],"funders":[],"has_content":{"grobid_xml":false,"pdf":false},"content_urls":null,"referenced_works_count":0,"referenced_works":[],"related_works":[],"abstract_inverted_index":{"As":[0],"large":[1],"language":[2,32],"models":[3,55,65,91,108,121],"(LLMs)":[4],"increasingly":[5],"interact":[6],"with":[7,134],"external":[8],"tools,":[9],"reward":[10,26,54,64,90,107],"modeling":[11],"for":[12,77,88,109],"tool":[13,72],"use":[14],"has":[15],"emerged":[16],"as":[17],"a":[18,85,104,131],"critical":[19],"yet":[20],"underexplored":[21],"area":[22],"of":[23,70,106,151],"research.":[24],"Existing":[25],"models,":[27],"trained":[28],"primarily":[29],"on":[30],"natural":[31],"outputs,":[33],"struggle":[34],"to":[35,51,114,130,141],"evaluate":[36,53],"tool-based":[37],"reasoning":[38],"and":[39,148],"execution.":[40],"To":[41],"quantify":[42],"this":[43,82],"gap,":[44],"we":[45],"introduce":[46,101],"FC-RewardBench,":[47],"the":[48,75],"first":[49],"benchmark":[50],"systematically":[52],"in":[56],"tool-calling":[57],"scenarios.":[58],"Our":[59],"analysis":[60],"shows":[61],"that":[62],"current":[63],"frequently":[66],"miss":[67],"key":[68],"signals":[69],"effective":[71,145],"use,":[73],"highlighting":[74],"need":[76],"domain-specific":[78],"modeling.":[79],"We":[80,100],"address":[81],"by":[83],"proposing":[84],"training":[86],"framework":[87],"outcome":[89],"using":[92],"data":[93,146],"synthesized":[94],"from":[95,112],"permissively":[96],"licensed,":[97],"open-weight":[98],"LLMs.":[99],"ToolRM":[102],"-":[103],"suite":[105],"tool-use":[110],"ranging":[111],"1.7B":[113],"14B":[115],"parameters.":[116],"Across":[117],"diverse":[118],"settings,":[119],"these":[120],"consistently":[122],"outperform":[123],"general-purpose":[124],"baselines.":[125],"Notably,":[126],"they":[127],"achieve":[128],"up":[129],"25%":[132],"improvement":[133],"Best-of-N":[135],"sampling,":[136],"while":[137],"also":[138],"improving":[139],"robustness":[140],"input":[142],"noise,":[143],"enabling":[144],"filtering,":[147],"supporting":[149],"RL-training":[150],"policy":[152],"models.":[153]},"counts_by_year":[],"updated_date":"2026-10-01T08:54:57.152042","created_date":"2025-10-12T00:00:00"}