{"id":"https://openalex.org/W4412945741","doi":"https://doi.org/10.18653/v1/2025.acl-long.257","title":"Mitigating Visual Forgetting via Take-along Visual Conditioning for Multi-modal Long CoT Reasoning","display_name":"Mitigating Visual Forgetting via Take-along Visual Conditioning for Multi-modal Long CoT Reasoning","publication_year":2025,"publication_date":"2025-01-01","ids":{"openalex":"https://openalex.org/W4412945741","doi":"https://doi.org/10.18653/v1/2025.acl-long.257"},"language":"en","primary_location":{"id":"doi:10.18653/v1/2025.acl-long.257","is_oa":true,"landing_page_url":"https://doi.org/10.18653/v1/2025.acl-long.257","pdf_url":"https://aclanthology.org/2025.acl-long.257.pdf","source":null,"license":"cc-by","license_id":"https://openalex.org/licenses/cc-by","version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"Proceedings of the 63rd Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers)","raw_type":"proceedings-article"},"type":"conference-paper","indexed_in":["crossref"],"open_access":{"is_oa":true,"oa_status":"gold","oa_url":"https://aclanthology.org/2025.acl-long.257.pdf","any_repository_has_fulltext":null},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5087952294","display_name":"Hailong Sun","orcid":"https://orcid.org/0000-0001-7654-5574"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Hai-Long Sun","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5016846734","display_name":"Zhun Sun","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Zhun Sun","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5080495327","display_name":"Houwen Peng","orcid":"https://orcid.org/0000-0001-8544-8952"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Houwen Peng","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"last","author":{"id":"https://openalex.org/A5065180062","display_name":"Han-Jia Ye","orcid":"https://orcid.org/0000-0003-1173-1880"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Han-Jia Ye","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":true,"cited_by_count":0,"citation_normalized_percentile":null,"cited_by_percentile_year":null,"biblio":{"volume":null,"issue":null,"first_page":"5158","last_page":"5171"},"is_retracted":false,"is_paratext":false,"is_xpac":false,"primary_topic":{"id":"https://openalex.org/T11714","display_name":"Multimodal Machine Learning Applications","score":0.9962999820709229,"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":0.9962999820709229,"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":0.9782999753952026,"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/T11596","display_name":"Constraint Satisfaction and Optimization","score":0.9345999956130981,"subfield":{"id":"https://openalex.org/subfields/1705","display_name":"Computer Networks and Communications"},"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/forgetting","display_name":"Forgetting","score":0.6568508744239807},{"id":"https://openalex.org/keywords/modal","display_name":"Modal","score":0.6264467239379883},{"id":"https://openalex.org/keywords/computer-science","display_name":"Computer science","score":0.6063814163208008},{"id":"https://openalex.org/keywords/conditioning","display_name":"Conditioning","score":0.5607995390892029},{"id":"https://openalex.org/keywords/artificial-intelligence","display_name":"Artificial intelligence","score":0.34604108333587646},{"id":"https://openalex.org/keywords/cognitive-psychology","display_name":"Cognitive psychology","score":0.3045373260974884},{"id":"https://openalex.org/keywords/psychology","display_name":"Psychology","score":0.2085966169834137},{"id":"https://openalex.org/keywords/mathematics","display_name":"Mathematics","score":0.061489641666412354}],"concepts":[{"id":"https://openalex.org/C7149132","wikidata":"https://www.wikidata.org/wiki/Q1377840","display_name":"Forgetting","level":2,"score":0.6568508744239807},{"id":"https://openalex.org/C71139939","wikidata":"https://www.wikidata.org/wiki/Q910194","display_name":"Modal","level":2,"score":0.6264467239379883},{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.6063814163208008},{"id":"https://openalex.org/C45262634","wikidata":"https://www.wikidata.org/wiki/Q5159291","display_name":"Conditioning","level":2,"score":0.5607995390892029},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.34604108333587646},{"id":"https://openalex.org/C180747234","wikidata":"https://www.wikidata.org/wiki/Q23373","display_name":"Cognitive psychology","level":1,"score":0.3045373260974884},{"id":"https://openalex.org/C15744967","wikidata":"https://www.wikidata.org/wiki/Q9418","display_name":"Psychology","level":0,"score":0.2085966169834137},{"id":"https://openalex.org/C33923547","wikidata":"https://www.wikidata.org/wiki/Q395","display_name":"Mathematics","level":0,"score":0.061489641666412354},{"id":"https://openalex.org/C185592680","wikidata":"https://www.wikidata.org/wiki/Q2329","display_name":"Chemistry","level":0,"score":0.0},{"id":"https://openalex.org/C105795698","wikidata":"https://www.wikidata.org/wiki/Q12483","display_name":"Statistics","level":1,"score":0.0},{"id":"https://openalex.org/C188027245","wikidata":"https://www.wikidata.org/wiki/Q750446","display_name":"Polymer chemistry","level":1,"score":0.0}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.18653/v1/2025.acl-long.257","is_oa":true,"landing_page_url":"https://doi.org/10.18653/v1/2025.acl-long.257","pdf_url":"https://aclanthology.org/2025.acl-long.257.pdf","source":null,"license":"cc-by","license_id":"https://openalex.org/licenses/cc-by","version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"Proceedings of the 63rd Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers)","raw_type":"proceedings-article"}],"best_oa_location":{"id":"doi:10.18653/v1/2025.acl-long.257","is_oa":true,"landing_page_url":"https://doi.org/10.18653/v1/2025.acl-long.257","pdf_url":"https://aclanthology.org/2025.acl-long.257.pdf","source":null,"license":"cc-by","license_id":"https://openalex.org/licenses/cc-by","version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"Proceedings of the 63rd Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers)","raw_type":"proceedings-article"},"sustainable_development_goals":[{"display_name":"Reduced inequalities","id":"https://metadata.un.org/sdg/10","score":0.46000000834465027}],"awards":[{"id":"https://openalex.org/G272683072","display_name":null,"funder_award_id":"62376118","funder_id":"https://openalex.org/F4320321001","funder_display_name":"National Natural Science Foundation of China"}],"funders":[{"id":"https://openalex.org/F4320311405","display_name":"Japan Agency for Medical Research and Development","ror":"https://ror.org/004rtk039"},{"id":"https://openalex.org/F4320316083","display_name":"Tencent","ror":"https://ror.org/00hhjss72"},{"id":"https://openalex.org/F4320321001","display_name":"National Natural Science Foundation of China","ror":"https://ror.org/01h0zpd94"}],"has_content":{"pdf":true,"grobid_xml":true},"content_urls":{"pdf":"https://content.openalex.org/works/W4412945741.pdf","grobid_xml":"https://content.openalex.org/works/W4412945741.grobid-xml"},"referenced_works_count":0,"referenced_works":[],"related_works":["https://openalex.org/W4289718052","https://openalex.org/W2164121020","https://openalex.org/W2145559838","https://openalex.org/W2905319430","https://openalex.org/W3116498279","https://openalex.org/W4287549553","https://openalex.org/W4310285384","https://openalex.org/W3183027292","https://openalex.org/W4248896073","https://openalex.org/W1964864628"],"abstract_inverted_index":{"Recent":[0],"advancements":[1],"in":[2,32,52,61,177],"Large":[3],"Language":[4],"Models":[5],"(LLMs)":[6],"have":[7],"demonstrated":[8],"enhanced":[9],"reasoning":[10,67,84,90,115,133,165,180],"capabilities,":[11],"evolving":[12],"from":[13,57],"Chainof-Thought":[14],"(CoT)":[15],"prompting":[16],"to":[17,45,63,131,149],"advanced,":[18],"product-oriented":[19],"solutions":[20],"like":[21],"OpenAI":[22],"o1.During":[23],"our":[24],"re-implementation":[25],"of":[26,175],"this":[27],"model,":[28],"we":[29,74,81,119],"noticed":[30],"that":[31,127],"multimodal":[33,179],"tasks":[34],"requiring":[35],"visual":[36,50,64,138,151],"input":[37,94,130],"(e.g.,":[38],"geometry":[39],"problems),":[40],"Multimodal":[41],"LLMs":[42],"(MLLMs)":[43],"struggle":[44],"maintain":[46],"focus":[47],"on":[48,103,160],"the":[49,83,89,93,108,113,145,150,154,173],"information,":[51],"other":[53],"words,":[54],"MLLMs":[55],"suffer":[56],"a":[58,99,125],"gradual":[59],"decline":[60],"attention":[62,148],"information":[65],"as":[66],"progresses,":[68],"causing":[69],"text-over-relied":[70],"outputs.To":[71],"investigate":[72],"this,":[73,118],"ablate":[75],"image":[76,95,129],"inputs":[77],"during":[78],"long-chain":[79],"reasoning.Concretely,":[80],"truncate":[82],"process":[85,91],"midway,":[86],"then":[87],"re-complete":[88],"with":[92],"removed.We":[96],"observe":[97],"only":[98],"2%":[100],"accuracy":[101],"drop":[102],"MathVista's":[104],"test-hard":[105],"subset,":[106],"revealing":[107],"model's":[109],"textual":[110],"outputs":[111],"dominate":[112],"following":[114],"process.Motivated":[116],"by":[117],"propose":[120],"Take-along":[121],"Visual":[122],"Conditioning":[123],"(TVC),":[124],"strategy":[126],"shifts":[128],"critical":[132],"stages":[134],"and":[135],"compresses":[136],"redundant":[137],"tokens":[139],"via":[140],"dynamic":[141],"pruning.This":[142],"methodology":[143],"helps":[144],"model":[146],"retain":[147],"components":[152],"throughout":[153],"reasoning.Our":[155],"approach":[156],"achieves":[157],"state-of-the-art":[158],"performance":[159],"average":[161],"across":[162],"five":[163],"mathematical":[164],"benchmarks":[166],"(+3.4":[167],"points":[168],"vs":[169],"previous":[170],"sota),":[171],"demonstrating":[172],"effectiveness":[174],"TVC":[176],"enhancing":[178],"systems.":[181]},"counts_by_year":[],"updated_date":"2026-07-14T23:27:15.235271","created_date":"2025-10-10T00:00:00"}