{"id":"https://openalex.org/W7204901457","doi":"https://doi.org/10.48550/arxiv.2608.30676","title":"MedAgent-R1: Faithfulness-Aware Reinforcement Learning for Evidence-Grounded Medical Reasoning","display_name":"MedAgent-R1: Faithfulness-Aware Reinforcement Learning for Evidence-Grounded Medical Reasoning","publication_year":2026,"publication_date":"2026-08-31","ids":{"openalex":"https://openalex.org/W7204901457","doi":"https://doi.org/10.48550/arxiv.2608.30676"},"language":"en","primary_location":{"id":"pmh:oai:arXiv.org:2608.30676","doi":null,"is_oa":true,"landing_page_url":"https://arxiv.org/abs/2608.30676","pdf_url":"https://arxiv.org/pdf/2608.30676","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","is_traditional":true,"indexed_in":["arxiv","datacite"],"open_access":{"is_oa":true,"oa_status":"green","oa_url":"https://arxiv.org/pdf/2608.30676","any_repository_has_fulltext":true},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5148848811","display_name":"Jiangwang Chen","orcid":null,"observed_orcids":[]},"institutions":[{"id":"https://openalex.org/I99065089","display_name":"Tsinghua University","ror":"https://ror.org/03cve4549","country_code":"CN","type":"education","lineage":["https://openalex.org/I99065089"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Chen, Jiangwang","raw_affiliation_strings":["Tsinghua University"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Tsinghua University","institution_ids":["https://openalex.org/I99065089"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5148876204","display_name":"Chenghao Zhang","orcid":null,"observed_orcids":[]},"institutions":[{"id":"https://openalex.org/I99065089","display_name":"Tsinghua University","ror":"https://ror.org/03cve4549","country_code":"CN","type":"education","lineage":["https://openalex.org/I99065089"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Zhang, Chenghao","raw_affiliation_strings":["Tsinghua University"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Tsinghua University","institution_ids":["https://openalex.org/I99065089"]}]},{"author_position":"last","author":{"id":"https://openalex.org/A5148891293","display_name":"Hengxing Cai","orcid":null,"observed_orcids":[]},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Cai, Hengxing","raw_affiliation_strings":["DP Technology * Equal contribution"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"DP Technology * Equal contribution","institution_ids":[]}]}],"institutions":[],"countries_distinct_count":1,"institutions_distinct_count":1,"corresponding_author_ids":[],"corresponding_institution_ids":[],"apc_list":null,"apc_paid":null,"fwci":null,"has_fulltext":true,"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/T11636","display_name":"Artificial Intelligence in Healthcare and Education","score":0.9560999870300293,"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"}},"topics":[{"id":"https://openalex.org/T11636","display_name":"Artificial Intelligence in Healthcare and Education","score":0.9560999870300293,"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"}},{"id":"https://openalex.org/T10286","display_name":"Information Retrieval and Search Behavior","score":0.012000000104308128,"subfield":{"id":"https://openalex.org/subfields/1710","display_name":"Information Systems"},"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/T13702","display_name":"Machine Learning in Healthcare","score":0.007600000128149986,"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/hallucinations","display_name":"hallucinations","score":0.5189999938011169},{"id":"https://openalex.org/keywords/faithfulness","display_name":"faithfulness","score":0.7369999885559082},{"id":"https://openalex.org/keywords/evidence-grounding","display_name":"evidence grounding","score":0.7900000214576721},{"id":"https://openalex.org/keywords/retrieval-augmented-generation","display_name":"retrieval-augmented generation","score":0.7789999842643738},{"id":"https://openalex.org/keywords/reinforcement-learning","display_name":"reinforcement learning","score":0.578000009059906},{"id":"https://openalex.org/keywords/reward-function-design","display_name":"reward function design","score":0.8180000185966492}],"concepts":[{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.6506999731063843},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.5637000203132629},{"id":"https://openalex.org/C97541855","wikidata":"https://www.wikidata.org/wiki/Q830687","display_name":"Reinforcement learning","level":2,"score":0.5044999718666077},{"id":"https://openalex.org/C17231256","wikidata":"https://www.wikidata.org/wiki/Q5156540","display_name":"Completeness (order theory)","level":2,"score":0.48069998621940613},{"id":"https://openalex.org/C119857082","wikidata":"https://www.wikidata.org/wiki/Q2539","display_name":"Machine learning","level":1,"score":0.42829999327659607},{"id":"https://openalex.org/C2778805511","wikidata":"https://www.wikidata.org/wiki/Q1713","display_name":"Citation","level":2,"score":0.4020000100135803},{"id":"https://openalex.org/C180747234","wikidata":"https://www.wikidata.org/wiki/Q23373","display_name":"Cognitive psychology","level":1,"score":0.3702000081539154},{"id":"https://openalex.org/C117251300","wikidata":"https://www.wikidata.org/wiki/Q1849855","display_name":"Parametric statistics","level":2,"score":0.3425000011920929},{"id":"https://openalex.org/C61797465","wikidata":"https://www.wikidata.org/wiki/Q1188986","display_name":"Term (time)","level":2,"score":0.33219999074935913},{"id":"https://openalex.org/C534262118","wikidata":"https://www.wikidata.org/wiki/Q177719","display_name":"Medical diagnosis","level":2,"score":0.3061000108718872},{"id":"https://openalex.org/C158154518","wikidata":"https://www.wikidata.org/wiki/Q7310970","display_name":"Relevance (law)","level":2,"score":0.2924000024795532},{"id":"https://openalex.org/C23123220","wikidata":"https://www.wikidata.org/wiki/Q816826","display_name":"Information retrieval","level":1,"score":0.29179999232292175},{"id":"https://openalex.org/C86037889","wikidata":"https://www.wikidata.org/wiki/Q4330127","display_name":"Learning to rank","level":3,"score":0.2903999984264374},{"id":"https://openalex.org/C15744967","wikidata":"https://www.wikidata.org/wiki/Q9418","display_name":"Psychology","level":0,"score":0.2888999879360199},{"id":"https://openalex.org/C2780451532","wikidata":"https://www.wikidata.org/wiki/Q759676","display_name":"Task (project management)","level":2,"score":0.28299999237060547},{"id":"https://openalex.org/C39920170","wikidata":"https://www.wikidata.org/wiki/Q693083","display_name":"Soundness","level":2,"score":0.28110000491142273},{"id":"https://openalex.org/C89611455","wikidata":"https://www.wikidata.org/wiki/Q6804646","display_name":"Mechanism (biology)","level":2,"score":0.2750999927520752},{"id":"https://openalex.org/C2780023022","wikidata":"https://www.wikidata.org/wiki/Q1338171","display_name":"Compensation (psychology)","level":2,"score":0.273499995470047},{"id":"https://openalex.org/C99844830","wikidata":"https://www.wikidata.org/wiki/Q102441924","display_name":"Scaling","level":2,"score":0.2727000117301941},{"id":"https://openalex.org/C108650721","wikidata":"https://www.wikidata.org/wiki/Q1783253","display_name":"Counterfactual thinking","level":2,"score":0.27160000801086426},{"id":"https://openalex.org/C50335755","wikidata":"https://www.wikidata.org/wiki/Q483247","display_name":"Phenomenon","level":2,"score":0.27140000462532043},{"id":"https://openalex.org/C2781181686","wikidata":"https://www.wikidata.org/wiki/Q4226068","display_name":"Coherence (philosophical gambling strategy)","level":2,"score":0.2700999975204468},{"id":"https://openalex.org/C125112378","wikidata":"https://www.wikidata.org/wiki/Q176640","display_name":"Randomness","level":2,"score":0.26589998602867126},{"id":"https://openalex.org/C2522767166","wikidata":"https://www.wikidata.org/wiki/Q2374463","display_name":"Data science","level":1,"score":0.2596000134944916},{"id":"https://openalex.org/C136389625","wikidata":"https://www.wikidata.org/wiki/Q334384","display_name":"Supervised learning","level":3,"score":0.258899986743927},{"id":"https://openalex.org/C18296254","wikidata":"https://www.wikidata.org/wiki/Q1395219","display_name":"Skepticism","level":2,"score":0.25209999084472656},{"id":"https://openalex.org/C2780801425","wikidata":"https://www.wikidata.org/wiki/Q5164392","display_name":"Construct (python library)","level":2,"score":0.25099998712539673}],"mesh":[],"locations_count":2,"locations":[{"id":"pmh:oai:arXiv.org:2608.30676","doi":null,"is_oa":true,"landing_page_url":"https://arxiv.org/abs/2608.30676","pdf_url":"https://arxiv.org/pdf/2608.30676","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.2608.30676","doi":"https://doi.org/10.48550/arxiv.2608.30676","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2608.30676","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":false,"raw_source_name":null,"raw_type":"Preprint"}],"best_oa_location":{"id":"pmh:oai:arXiv.org:2608.30676","doi":null,"is_oa":true,"landing_page_url":"https://arxiv.org/abs/2608.30676","pdf_url":"https://arxiv.org/pdf/2608.30676","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":[],"study_designs":[],"awards":[],"funders":[],"has_content":{"grobid_xml":true,"pdf":true},"content_urls":{"pdf":"https://content.openalex.org/works/W7204901457.pdf","grobid_xml":"https://content.openalex.org/works/W7204901457.grobid-xml"},"referenced_works_count":0,"referenced_works":[],"related_works":["https://openalex.org/W7167270975","https://openalex.org/W7166843029","https://openalex.org/W7170275858","https://openalex.org/W7164988430","https://openalex.org/W7171929337","https://openalex.org/W7166819041","https://openalex.org/W7168035881","https://openalex.org/W7213672938","https://openalex.org/W7127638109","https://openalex.org/W7168173065"],"abstract_inverted_index":{"When":[0],"medical":[1],"AI":[2],"systems":[3],"hallucinate":[4],"clinical":[5],"reasoning,":[6],"the":[7,101,168],"consequences":[8],"extend":[9],"beyond":[10],"incorrect":[11],"answers:":[12],"fabricated":[13],"justifications":[14,40],"that":[15,41,130,196],"superficially":[16],"reference":[17],"retrieved":[18],"evidence":[19,117,151],"can":[20,43],"mislead":[21],"clinicians":[22,42],"into":[23],"unsafe":[24],"treatment":[25],"decisions.":[26],"Medical":[27],"reasoning":[28],"agents":[29],"must":[30],"therefore":[31],"produce":[32],"not":[33,203],"only":[34],"correct":[35],"answers":[36],"but":[37,82],"also":[38],"faithful":[39],"verify":[44],"against":[45],"cited":[46],"evidence.":[47],"We":[48,104],"identify":[49],"a":[50,65,108,120],"systematic":[51],"failure":[52],"mode":[53],"in":[54,192],"RL-trained":[55],"retrieval":[56,125,171],"agents:":[57],"outcome-only":[58],"rewards":[59],"improve":[60],"accuracy":[61,95,112],"while":[62,157,188],"degrading":[63],"faithfulness,":[64],"phenomenon":[66],"we":[67],"term":[68],"confident":[69],"hallucination.":[70],"The":[71,138],"agent":[72],"learns":[73],"to":[74,91,135,147,155],"answer":[75],"from":[76,89,145,153],"parametric":[77],"memory":[78],"and":[79,127,149],"backfill":[80],"plausible":[81],"unsupported":[83],"justifications;":[84],"citation":[85,143],"fabrication":[86,144],"rates":[87],"rise":[88],"16.5%":[90],"31.8%":[92,146],"even":[93],"as":[94],"improves":[96],"by":[97,124,205],"5":[98],"points":[99],"over":[100],"supervised":[102],"baseline.":[103],"address":[105],"this":[106],"with":[107,161],"faithfulness-gated":[109],"reward":[110],"design:":[111],"credit":[113],"is":[114],"conditioned":[115],"on":[116,164,176],"grounding":[118],"via":[119],"hard":[121],"gate,":[122],"complemented":[123],"validity":[126],"conciseness":[128],"signals":[129],"close":[131],"exploitation":[132],"paths":[133],"unique":[134],"agentic":[136,170],"retrieval.":[137],"resulting":[139],"system,":[140],"MedAgent-R1,":[141],"reduces":[142],"4.7%":[148],"raises":[150],"completeness":[152],"58.7":[154],"82.6":[156],"maintaining":[158],"75.1%":[159],"accuracy,":[160,194],"13.2-point":[162],"gains":[163,202],"HealthBench":[165],"Safety.":[166],"Under":[167],"same":[169],"setup,":[172],"MedAgent-R1":[173],"outscores":[174],"GPT-4o":[175,191],"faithfulness-specific":[177],"dimensions":[178],"(Factual":[179],"Support":[180],"4.55":[181],"vs.":[182,186],"4.25;":[183],"Overclaiming":[184],"4.40":[185],"4.15)":[187],"remaining":[189],"below":[190],"overall":[193],"suggesting":[195],"explicit":[197],"faithfulness":[198],"training":[199],"yields":[200],"evidence-grounding":[201],"achieved":[204],"scaling":[206],"alone.":[207]},"counts_by_year":[],"updated_date":"2026-11-04T10:54:05.112534","created_date":"2026-09-02T00:00:00"}