{"id":"https://openalex.org/W7160836721","doi":"https://doi.org/10.48550/arxiv.2605.07477","title":"ReasonEdit: Towards Interpretable Image Editing Evaluation via Reinforcement Learning","display_name":"ReasonEdit: Towards Interpretable Image Editing Evaluation via Reinforcement Learning","publication_year":2026,"publication_date":"2026-05-08","ids":{"openalex":"https://openalex.org/W7160836721","doi":"https://doi.org/10.48550/arxiv.2605.07477"},"language":null,"primary_location":{"id":"doi:10.48550/arxiv.2605.07477","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2605.07477","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":null,"license_id":null,"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.2605.07477","any_repository_has_fulltext":true},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5135883476","display_name":"Honghua Chen","orcid":"https://orcid.org/0000-0002-3455-5333"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Chen, Honghua","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5135905001","display_name":"Zitong Xu","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Xu, Zitong","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5135869321","display_name":"Huiyu Duan","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Duan, Huiyu","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5135900991","display_name":"Xinyun Zhang","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Zhang, Xinyun","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5135879920","display_name":"Xiongkuo Min","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Min, Xiongkuo","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"last","author":{"id":"https://openalex.org/A5135833169","display_name":"Guangtao Zhai","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Zhai, Guangtao","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/T10775","display_name":"Generative Adversarial Networks and Image Synthesis","score":0.5386999845504761,"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"}},"topics":[{"id":"https://openalex.org/T10775","display_name":"Generative Adversarial Networks and Image Synthesis","score":0.5386999845504761,"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"}},{"id":"https://openalex.org/T11714","display_name":"Multimodal Machine Learning Applications","score":0.0625,"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"}},{"id":"https://openalex.org/T11605","display_name":"Visual Attention and Saliency Detection","score":0.04179999977350235,"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/generalization","display_name":"Generalization","score":0.5867000222206116},{"id":"https://openalex.org/keywords/reinforcement-learning","display_name":"Reinforcement learning","score":0.5123999714851379},{"id":"https://openalex.org/keywords/image","display_name":"Image (mathematics)","score":0.5054000020027161},{"id":"https://openalex.org/keywords/interpretation","display_name":"Interpretation (philosophy)","score":0.47690001130104065},{"id":"https://openalex.org/keywords/trustworthiness","display_name":"Trustworthiness","score":0.45500001311302185},{"id":"https://openalex.org/keywords/code","display_name":"Code (set theory)","score":0.4068000018596649}],"concepts":[{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.7269999980926514},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.6614999771118164},{"id":"https://openalex.org/C119857082","wikidata":"https://www.wikidata.org/wiki/Q2539","display_name":"Machine learning","level":1,"score":0.5892999768257141},{"id":"https://openalex.org/C177148314","wikidata":"https://www.wikidata.org/wiki/Q170084","display_name":"Generalization","level":2,"score":0.5867000222206116},{"id":"https://openalex.org/C97541855","wikidata":"https://www.wikidata.org/wiki/Q830687","display_name":"Reinforcement learning","level":2,"score":0.5123999714851379},{"id":"https://openalex.org/C115961682","wikidata":"https://www.wikidata.org/wiki/Q860623","display_name":"Image (mathematics)","level":2,"score":0.5054000020027161},{"id":"https://openalex.org/C527412718","wikidata":"https://www.wikidata.org/wiki/Q855395","display_name":"Interpretation (philosophy)","level":2,"score":0.47690001130104065},{"id":"https://openalex.org/C153701036","wikidata":"https://www.wikidata.org/wiki/Q659974","display_name":"Trustworthiness","level":2,"score":0.45500001311302185},{"id":"https://openalex.org/C2776760102","wikidata":"https://www.wikidata.org/wiki/Q5139990","display_name":"Code (set theory)","level":3,"score":0.4068000018596649},{"id":"https://openalex.org/C75294576","wikidata":"https://www.wikidata.org/wiki/Q5165192","display_name":"Contextual image classification","level":3,"score":0.33820000290870667},{"id":"https://openalex.org/C2780801425","wikidata":"https://www.wikidata.org/wiki/Q5164392","display_name":"Construct (python library)","level":2,"score":0.27799999713897705},{"id":"https://openalex.org/C204321447","wikidata":"https://www.wikidata.org/wiki/Q30642","display_name":"Natural language processing","level":1,"score":0.2689000070095062},{"id":"https://openalex.org/C49937458","wikidata":"https://www.wikidata.org/wiki/Q2599292","display_name":"Probabilistic logic","level":2,"score":0.2653999924659729},{"id":"https://openalex.org/C153180895","wikidata":"https://www.wikidata.org/wiki/Q7148389","display_name":"Pattern recognition (psychology)","level":2,"score":0.2630000114440918}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.48550/arxiv.2605.07477","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2605.07477","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":null,"license_id":null,"version":null,"is_accepted":false,"is_published":null,"raw_source_name":null,"raw_type":"Preprint"}],"best_oa_location":{"id":"doi:10.48550/arxiv.2605.07477","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2605.07477","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":null,"license_id":null,"version":null,"is_accepted":false,"is_published":false,"raw_source_name":null,"raw_type":"Preprint"},"sustainable_development_goals":[{"display_name":"Peace, Justice and strong institutions","id":"https://metadata.un.org/sdg/16","score":0.46537479758262634}],"awards":[],"funders":[],"has_content":{"grobid_xml":false,"pdf":false},"content_urls":null,"referenced_works_count":0,"referenced_works":[],"related_works":[],"abstract_inverted_index":{"Recent":[0],"text-guided":[1],"image":[2,123,188],"editing":[3],"(TIE)":[4],"models":[5,58],"have":[6,29],"achieved":[7],"remarkable":[8],"progress,":[9],"however,":[10],"many":[11],"edited":[12,76],"results":[13],"still":[14],"suffer":[15],"from":[16,46,136],"artifacts,":[17],"unintended":[18],"modifications,":[19],"and":[20,26,39,55,96,138,163,184],"suboptimal":[21],"aesthetics.":[22],"Although":[23],"several":[24],"benchmarks":[25],"evaluation":[27,150,179],"methods":[28],"been":[30],"proposed,":[31],"most":[32],"existing":[33],"approaches":[34],"rely":[35],"on":[36],"scalar":[37],"scores":[38],"lack":[40],"interpretability.":[41],"This":[42],"limitation":[43],"largely":[44],"stems":[45],"the":[47,70,139],"absence":[48],"of":[49,93,175],"high-quality":[50,177],"interpretation":[51],"datasets":[52],"for":[53,118,187],"TIE":[54],"effective":[56],"reward":[57,111,133],"to":[59,114,146],"train":[60],"interpretable":[61,120,149,178],"evaluators.":[62],"To":[63],"address":[64],"these":[65,89],"challenges,":[66],"we":[67,102,126],"introduce":[68],"ReasonEdit-22K,":[69],"first":[71],"dataset":[72],"that":[73,155],"combines":[74],"22K":[75],"images":[77],"with":[78,84,160],"113K":[79],"Chain-of-Thought":[80],"(CoT)":[81],"samples,":[82],"along":[83],"1.3M":[85],"human":[86,161],"judgments":[87],"assessing":[88],"interpretations":[90],"in":[91,122],"terms":[92],"logicality,":[94],"accuracy,":[95],"usefulness.":[97],"Building":[98],"upon":[99],"this":[100],"dataset,":[101],"propose":[103],"RE-Reward,":[104],"a":[105],"multimodal":[106],"large":[107],"language":[108],"model":[109,112],"(MLLM)-based":[110],"designed":[113],"provide":[115],"human-aligned":[116],"feedback":[117],"evaluating":[119],"reasoning":[121],"editing.":[124,189],"Furthermore,":[125],"develop":[127],"ReasonEdit,":[128],"which":[129],"is":[130,173,192],"trained":[131],"using":[132],"signals":[134],"derived":[135],"RE-Reward":[137],"Group":[140],"Relative":[141],"Policy":[142],"Optimization":[143],"(GRPO)":[144],"algorithm":[145],"learn":[147],"an":[148],"model.":[151],"Extensive":[152],"experiments":[153],"demonstrate":[154],"ReasonEdit":[156],"achieves":[157],"superior":[158],"alignment":[159],"preferences":[162],"exhibits":[164],"strong":[165],"generalization":[166],"across":[167],"public":[168],"benchmarks.":[169],"In":[170],"addition,":[171],"it":[172],"capable":[174],"generating":[176],"text,":[180],"enabling":[181],"more":[182],"transparent":[183],"trustworthy":[185],"assessment":[186],"The":[190],"code":[191],"available":[193],"at":[194],"https://github.com/IntMeGroup/ReasonEdit.":[195]},"counts_by_year":[],"updated_date":"2026-07-01T06:00:48.157686","created_date":"2026-05-12T00:00:00"}