{"id":"https://openalex.org/W7204173890","doi":"https://doi.org/10.48550/arxiv.2608.22429","title":"Think with Structured Grounding: Perceptual Reinforcement Learning for Chart and Visual-Tabular Understanding","display_name":"Think with Structured Grounding: Perceptual Reinforcement Learning for Chart and Visual-Tabular Understanding","publication_year":2026,"publication_date":"2026-08-23","ids":{"openalex":"https://openalex.org/W7204173890","doi":"https://doi.org/10.48550/arxiv.2608.22429"},"language":"en","primary_location":{"id":"pmh:oai:arXiv.org:2608.22429","doi":null,"is_oa":true,"landing_page_url":"https://arxiv.org/abs/2608.22429","pdf_url":"https://arxiv.org/pdf/2608.22429","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/2608.22429","any_repository_has_fulltext":true},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5148098120","display_name":"Changjiang Jiang","orcid":null,"observed_orcids":[]},"institutions":[{"id":"https://openalex.org/I4210113480","display_name":"Mohamed bin Zayed University of Artificial Intelligence","ror":"https://ror.org/0258gkt32","country_code":"AE","type":"education","lineage":["https://openalex.org/I4210113480"]}],"countries":["AE"],"is_corresponding":false,"raw_author_name":"Jiang, Changjiang","raw_affiliation_strings":["Mohamed bin Zayed University of Artificial Intelligence"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Mohamed bin Zayed University of Artificial Intelligence","institution_ids":["https://openalex.org/I4210113480"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5098933159","display_name":"Qiannian ZhAO","orcid":"https://orcid.org/0009-0000-5385-8041","observed_orcids":["https://orcid.org/0009-0000-5385-8041"]},"institutions":[{"id":"https://openalex.org/I4210113480","display_name":"Mohamed bin Zayed University of Artificial Intelligence","ror":"https://ror.org/0258gkt32","country_code":"AE","type":"education","lineage":["https://openalex.org/I4210113480"]}],"countries":["AE"],"is_corresponding":false,"raw_author_name":"Zhao, Qiannian","raw_affiliation_strings":["Mohamed bin Zayed University of Artificial Intelligence"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Mohamed bin Zayed University of Artificial Intelligence","institution_ids":["https://openalex.org/I4210113480"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5126633090","display_name":"Lei Xin","orcid":null,"observed_orcids":[]},"institutions":[{"id":"https://openalex.org/I4210113480","display_name":"Mohamed bin Zayed University of Artificial Intelligence","ror":"https://ror.org/0258gkt32","country_code":"AE","type":"education","lineage":["https://openalex.org/I4210113480"]}],"countries":["AE"],"is_corresponding":false,"raw_author_name":"Xin, Lei","raw_affiliation_strings":["Mohamed bin Zayed University of Artificial Intelligence"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Mohamed bin Zayed University of Artificial Intelligence","institution_ids":["https://openalex.org/I4210113480"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5126398314","display_name":"Jinxiang Xie","orcid":null,"observed_orcids":[]},"institutions":[{"id":"https://openalex.org/I881766915","display_name":"Nanjing University","ror":"https://ror.org/01rxvg760","country_code":"CN","type":"education","lineage":["https://openalex.org/I881766915"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Xie, Jinxiang","raw_affiliation_strings":["Nanjing University"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Nanjing University","institution_ids":["https://openalex.org/I881766915"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5139740785","display_name":"Preslav Nakov","orcid":null,"observed_orcids":[]},"institutions":[{"id":"https://openalex.org/I4210113480","display_name":"Mohamed bin Zayed University of Artificial Intelligence","ror":"https://ror.org/0258gkt32","country_code":"AE","type":"education","lineage":["https://openalex.org/I4210113480"]}],"countries":["AE"],"is_corresponding":false,"raw_author_name":"Nakov, Preslav","raw_affiliation_strings":["Mohamed bin Zayed University of Artificial Intelligence"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Mohamed bin Zayed University of Artificial Intelligence","institution_ids":["https://openalex.org/I4210113480"]}]},{"author_position":"last","author":{"id":"https://openalex.org/A5151408937","display_name":"Zhuohan Xie","orcid":null,"observed_orcids":[]},"institutions":[{"id":"https://openalex.org/I4210113480","display_name":"Mohamed bin Zayed University of Artificial Intelligence","ror":"https://ror.org/0258gkt32","country_code":"AE","type":"education","lineage":["https://openalex.org/I4210113480"]}],"countries":["AE"],"is_corresponding":false,"raw_author_name":"Xie, Zhuohan","raw_affiliation_strings":["Mohamed bin Zayed University of Artificial Intelligence"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Mohamed bin Zayed University of Artificial Intelligence","institution_ids":["https://openalex.org/I4210113480"]}]}],"institutions":[],"countries_distinct_count":2,"institutions_distinct_count":2,"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/T11714","display_name":"Multimodal Machine Learning Applications","score":1.0,"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/T11714","display_name":"Multimodal Machine Learning Applications","score":1.0,"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/T10036","display_name":"Advanced Neural Network Applications","score":0.0,"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/T10601","display_name":"Handwritten Text Recognition Techniques","score":0.0,"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/multimodal-large-language-models","display_name":"multimodal large language models","score":0.9350000023841858},{"id":"https://openalex.org/keywords/chart-understanding","display_name":"chart understanding","score":0.656000018119812},{"id":"https://openalex.org/keywords/reinforcement-learning-from-human-feedback","display_name":"reinforcement learning from human feedback","score":0.7760000228881836},{"id":"https://openalex.org/keywords/inference-latency","display_name":"inference latency","score":0.4490000009536743},{"id":"https://openalex.org/keywords/visual-question-answering","display_name":"visual question answering","score":0.6610000133514404},{"id":"https://openalex.org/keywords/knowledge-distillation","display_name":"knowledge distillation","score":0.5019999742507935}],"concepts":[{"id":"https://openalex.org/C2776214188","wikidata":"https://www.wikidata.org/wiki/Q408386","display_name":"Inference","level":2,"score":0.7124999761581421},{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.7052000164985657},{"id":"https://openalex.org/C2777508537","wikidata":"https://www.wikidata.org/wiki/Q7936620","display_name":"Visual reasoning","level":2,"score":0.6288999915122986},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.5952000021934509},{"id":"https://openalex.org/C26760741","wikidata":"https://www.wikidata.org/wiki/Q160402","display_name":"Perception","level":2,"score":0.557200014591217},{"id":"https://openalex.org/C97541855","wikidata":"https://www.wikidata.org/wiki/Q830687","display_name":"Reinforcement learning","level":2,"score":0.5029000043869019},{"id":"https://openalex.org/C43521106","wikidata":"https://www.wikidata.org/wiki/Q2165493","display_name":"Pipeline (software)","level":2,"score":0.4864000082015991},{"id":"https://openalex.org/C190812933","wikidata":"https://www.wikidata.org/wiki/Q28923","display_name":"Chart","level":2,"score":0.4740999937057495},{"id":"https://openalex.org/C98045186","wikidata":"https://www.wikidata.org/wiki/Q205663","display_name":"Process (computing)","level":2,"score":0.4691999852657318},{"id":"https://openalex.org/C26517878","wikidata":"https://www.wikidata.org/wiki/Q228039","display_name":"Key (lock)","level":2,"score":0.40790000557899475},{"id":"https://openalex.org/C119857082","wikidata":"https://www.wikidata.org/wiki/Q2539","display_name":"Machine learning","level":1,"score":0.4050000011920929},{"id":"https://openalex.org/C178253425","wikidata":"https://www.wikidata.org/wiki/Q162668","display_name":"Visual perception","level":3,"score":0.3885999917984009},{"id":"https://openalex.org/C22367795","wikidata":"https://www.wikidata.org/wiki/Q7625208","display_name":"Structured prediction","level":2,"score":0.3580000102519989},{"id":"https://openalex.org/C107457646","wikidata":"https://www.wikidata.org/wiki/Q207434","display_name":"Human\u2013computer interaction","level":1,"score":0.3555000126361847},{"id":"https://openalex.org/C125411270","wikidata":"https://www.wikidata.org/wiki/Q18653","display_name":"Encoding (memory)","level":2,"score":0.35530000925064087},{"id":"https://openalex.org/C36464697","wikidata":"https://www.wikidata.org/wiki/Q451553","display_name":"Visualization","level":2,"score":0.3425999879837036},{"id":"https://openalex.org/C75553542","wikidata":"https://www.wikidata.org/wiki/Q178161","display_name":"A priori and a posteriori","level":2,"score":0.32440000772476196},{"id":"https://openalex.org/C155911833","wikidata":"https://www.wikidata.org/wiki/Q3817354","display_name":"Spatial intelligence","level":2,"score":0.31470000743865967},{"id":"https://openalex.org/C204321447","wikidata":"https://www.wikidata.org/wiki/Q30642","display_name":"Natural language processing","level":1,"score":0.3050999939441681},{"id":"https://openalex.org/C194995250","wikidata":"https://www.wikidata.org/wiki/Q531136","display_name":"Affordance","level":2,"score":0.29750001430511475},{"id":"https://openalex.org/C188147891","wikidata":"https://www.wikidata.org/wiki/Q147638","display_name":"Cognitive science","level":1,"score":0.2939000129699707},{"id":"https://openalex.org/C89288958","wikidata":"https://www.wikidata.org/wiki/Q7301504","display_name":"Reasoning system","level":2,"score":0.2921000123023987},{"id":"https://openalex.org/C82876162","wikidata":"https://www.wikidata.org/wiki/Q17096504","display_name":"Latency (audio)","level":2,"score":0.2809999883174896},{"id":"https://openalex.org/C195344581","wikidata":"https://www.wikidata.org/wiki/Q2555318","display_name":"Automated reasoning","level":2,"score":0.27059999108314514},{"id":"https://openalex.org/C2779458634","wikidata":"https://www.wikidata.org/wiki/Q24963715","display_name":"Debiasing","level":2,"score":0.2605000138282776},{"id":"https://openalex.org/C115901376","wikidata":"https://www.wikidata.org/wiki/Q184199","display_name":"Automation","level":2,"score":0.25679999589920044},{"id":"https://openalex.org/C12713177","wikidata":"https://www.wikidata.org/wiki/Q1900281","display_name":"Perspective (graphical)","level":2,"score":0.25540000200271606},{"id":"https://openalex.org/C166088908","wikidata":"https://www.wikidata.org/wiki/Q308495","display_name":"Abductive reasoning","level":2,"score":0.251800000667572}],"mesh":[],"locations_count":2,"locations":[{"id":"pmh:oai:arXiv.org:2608.22429","doi":null,"is_oa":true,"landing_page_url":"https://arxiv.org/abs/2608.22429","pdf_url":"https://arxiv.org/pdf/2608.22429","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.2608.22429","doi":"https://doi.org/10.48550/arxiv.2608.22429","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2608.22429","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:2608.22429","doi":null,"is_oa":true,"landing_page_url":"https://arxiv.org/abs/2608.22429","pdf_url":"https://arxiv.org/pdf/2608.22429","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":{"pdf":true,"grobid_xml":true},"content_urls":{"pdf":"https://content.openalex.org/works/W7204173890.pdf","grobid_xml":"https://content.openalex.org/works/W7204173890.grobid-xml"},"referenced_works_count":0,"referenced_works":[],"related_works":["https://openalex.org/W7163032927","https://openalex.org/W4415193793","https://openalex.org/W7154227658","https://openalex.org/W4416445907","https://openalex.org/W7133331446","https://openalex.org/W4405882635","https://openalex.org/W4415110418","https://openalex.org/W4393335917","https://openalex.org/W4406273429","https://openalex.org/W4417449769"],"abstract_inverted_index":{"Multimodal":[0],"Large":[1],"Language":[2],"Models":[3],"(MLLMs)":[4],"capable":[5],"of":[6,97,153],"thinking":[7],"with":[8,60,73,166,216],"images":[9],"often":[10],"rely":[11],"on":[12],"external":[13,55],"tools":[14],"for":[15],"fine-grained":[16,62,79,218],"perception.":[17],"However,":[18],"this":[19],"reliance":[20],"introduces":[21],"significant":[22],"inference":[23,206],"latency":[24,207],"and":[25,37,43,100,123,174,177,212,221],"fails":[26],"to":[27,84,115,129,170],"effectively":[28],"resolve":[29],"the":[30,91,95,146],"spatial-structural":[31],"gap-a":[32],"fundamental":[33],"challenge":[34],"in":[35],"text-dense":[36],"structurally":[38],"relational":[39],"visuals":[40],"(e.g.,":[41],"charts":[42],"visual":[44,63,132],"tables)":[45],"where":[46],"strict":[47],"relative":[48],"spatial":[49],"arrangements":[50],"bind":[51],"textual":[52],"elements.":[53],"Without":[54],"tools,":[56],"standard":[57],"MLLMs":[58],"struggle":[59],"such":[61],"reasoning":[64,99,173,223],"tasks.":[65],"To":[66],"address":[67],"these":[68],"issues,":[69],"we":[70,111],"propose":[71],"Think":[72],"Structured":[74],"Grounding":[75],"(TwSG),":[76],"a":[77,103,126,157,179,186],"novel":[78,187],"image":[80],"perception":[81],"framework":[82],"designed":[83],"internalize":[85],"complex":[86,172],"images's":[87],"tool-use":[88],"capabilities":[89],"within":[90],"model.":[92],"TwSG":[93,204],"distills":[94],"benefits":[96],"multi-step":[98],"micro-cropping":[101],"into":[102,145],"single":[104],"efficient":[105],"forward":[106],"pass":[107],"during":[108],"inference.":[109],"Specifically,":[110],"use":[112],"an":[113],"MLLM":[114,200],"identify":[116],"key":[117],"regions":[118],"guided":[119],"by":[120,185],"ground-truth":[121],"answers,":[122],"then":[124],"prompt":[125],"teacher":[127],"model":[128],"generate":[130],"high-quality":[131],"question-answering":[133],"(VQA)":[134],"data.":[135],"These":[136],"fine-grained,":[137],"region-based":[138],"supervisory":[139],"signals":[140],"are":[141],"subsequently":[142],"distilled":[143],"back":[144],"full-image":[147],"representation.":[148],"Our":[149],"training":[150],"pipeline":[151],"consists":[152],"two":[154],"stages:":[155],"(1)":[156],"cold-start":[158],"supervised":[159],"fine-tuning":[160,181],"(SFT)":[161],"phase":[162,183],"using":[163],"multi-turn":[164],"data":[165],"focused":[167],"area":[168],"descriptions":[169],"foster":[171],"error":[175],"recovery;":[176],"(2)":[178],"reinforcement":[180],"(RFT)":[182],"driven":[184],"process":[188],"reward":[189],"mechanism,":[190],"TL-GRPO,":[191],"which":[192],"encourages":[193],"strategic":[194],"reasoning.":[195],"Extensive":[196],"experiments":[197],"across":[198],"various":[199],"architectures":[201],"demonstrate":[202],"that":[203],"reduces":[205],"while":[208],"substantially":[209],"improving":[210],"accuracy":[211],"robustness,":[213],"endowing":[214],"models":[215],"native":[217],"region":[219],"description":[220],"flexible":[222],"capabilities.":[224]},"counts_by_year":[],"updated_date":"2026-10-28T10:33:24.094632","created_date":"2026-08-26T00:00:00"}