{"id":"https://openalex.org/W7166802427","doi":"https://doi.org/10.18653/v1/2026.findings-acl.1044","title":"Beyond Modality Collapse: Taming Guided Modality Entropy for Omni-modal Emotion Reasoning","display_name":"Beyond Modality Collapse: Taming Guided Modality Entropy for Omni-modal Emotion Reasoning","publication_year":2026,"publication_date":"2026-01-01","ids":{"openalex":"https://openalex.org/W7166802427","doi":"https://doi.org/10.18653/v1/2026.findings-acl.1044"},"language":null,"primary_location":{"id":"doi:10.18653/v1/2026.findings-acl.1044","is_oa":true,"landing_page_url":"https://doi.org/10.18653/v1/2026.findings-acl.1044","pdf_url":"https://aclanthology.org/2026.findings-acl.1044.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":"Findings of the Association for Computational Linguistics: ACL 2026","raw_type":"proceedings-article"},"type":"conference-paper","indexed_in":["crossref"],"open_access":{"is_oa":true,"oa_status":"gold","oa_url":"https://aclanthology.org/2026.findings-acl.1044.pdf","any_repository_has_fulltext":null},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5134100014","display_name":"Xian Zhao","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Xian Zhao","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5139770539","display_name":"Rui Hu","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Rui Hu","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5139756448","display_name":"Yuxiang Zhang","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Yuxiang Zhang","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5134142365","display_name":"Delai Qiu","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Delai Qiu","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5139817815","display_name":"Yining Wang","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Yining Wang","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5139809697","display_name":"Shengping Liu","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Shengping Liu","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5139848682","display_name":"Jian Yu","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Jian Yu","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"last","author":{"id":"https://openalex.org/A5139796076","display_name":"Jitao Sang","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Jitao Sang","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":0.0,"has_fulltext":true,"cited_by_count":0,"citation_normalized_percentile":{"value":0.79170656,"is_in_top_1_percent":false,"is_in_top_10_percent":false},"cited_by_percentile_year":null,"biblio":{"volume":null,"issue":null,"first_page":"20830","last_page":"20845"},"is_retracted":false,"is_paratext":false,"is_xpac":false,"primary_topic":{"id":"https://openalex.org/T12031","display_name":"Speech and dialogue systems","score":0.12210000306367874,"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/T12031","display_name":"Speech and dialogue systems","score":0.12210000306367874,"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/T11010","display_name":"Logic, Reasoning, and Knowledge","score":0.08139999955892563,"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/T11431","display_name":"Action Observation and Synchronization","score":0.07859999686479568,"subfield":{"id":"https://openalex.org/subfields/3207","display_name":"Social Psychology"},"field":{"id":"https://openalex.org/fields/32","display_name":"Psychology"},"domain":{"id":"https://openalex.org/domains/2","display_name":"Social Sciences"}}],"keywords":[{"id":"https://openalex.org/keywords/modality","display_name":"Modality (human\u2013computer interaction)","score":0.5827999711036682},{"id":"https://openalex.org/keywords/entropy","display_name":"Entropy (arrow of time)","score":0.4440000057220459},{"id":"https://openalex.org/keywords/component","display_name":"Component (thermodynamics)","score":0.24539999663829803}],"concepts":[{"id":"https://openalex.org/C2780226545","wikidata":"https://www.wikidata.org/wiki/Q6888030","display_name":"Modality (human\u2013computer interaction)","level":2,"score":0.5827999711036682},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.5449000000953674},{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.5343999862670898},{"id":"https://openalex.org/C106301342","wikidata":"https://www.wikidata.org/wiki/Q4117933","display_name":"Entropy (arrow of time)","level":2,"score":0.4440000057220459},{"id":"https://openalex.org/C15744967","wikidata":"https://www.wikidata.org/wiki/Q9418","display_name":"Psychology","level":0,"score":0.367900013923645},{"id":"https://openalex.org/C204321447","wikidata":"https://www.wikidata.org/wiki/Q30642","display_name":"Natural language processing","level":1,"score":0.33079999685287476},{"id":"https://openalex.org/C180747234","wikidata":"https://www.wikidata.org/wiki/Q23373","display_name":"Cognitive psychology","level":1,"score":0.287200003862381},{"id":"https://openalex.org/C188147891","wikidata":"https://www.wikidata.org/wiki/Q147638","display_name":"Cognitive science","level":1,"score":0.2599000036716461},{"id":"https://openalex.org/C168167062","wikidata":"https://www.wikidata.org/wiki/Q1117970","display_name":"Component (thermodynamics)","level":2,"score":0.24539999663829803},{"id":"https://openalex.org/C31972630","wikidata":"https://www.wikidata.org/wiki/Q844240","display_name":"Computer vision","level":1,"score":0.22529999911785126}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.18653/v1/2026.findings-acl.1044","is_oa":true,"landing_page_url":"https://doi.org/10.18653/v1/2026.findings-acl.1044","pdf_url":"https://aclanthology.org/2026.findings-acl.1044.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":"Findings of the Association for Computational Linguistics: ACL 2026","raw_type":"proceedings-article"}],"best_oa_location":{"id":"doi:10.18653/v1/2026.findings-acl.1044","is_oa":true,"landing_page_url":"https://doi.org/10.18653/v1/2026.findings-acl.1044","pdf_url":"https://aclanthology.org/2026.findings-acl.1044.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":"Findings of the Association for Computational Linguistics: ACL 2026","raw_type":"proceedings-article"},"sustainable_development_goals":[],"awards":[],"funders":[{"id":"https://openalex.org/F4320321001","display_name":"National Natural Science Foundation of China","ror":"https://ror.org/01h0zpd94"}],"has_content":{"grobid_xml":true,"pdf":true},"content_urls":{"pdf":"https://content.openalex.org/works/W7166802427.pdf","grobid_xml":"https://content.openalex.org/works/W7166802427.grobid-xml"},"referenced_works_count":0,"referenced_works":[],"related_works":[],"abstract_inverted_index":{"Omni-modal":[0],"Large":[1],"Language":[2],"Models":[3],"(OLLMs)":[4],"excel":[5],"in":[6],"diverse":[7],"tasks":[8],"but":[9],"struggle":[10],"with":[11,105],"complex":[12,155],"emotional":[13],"reasoning,":[14],"which":[15,83],"requires":[16],"integrating":[17],"textual,":[18],"visual,":[19],"and":[20,72,99,128],"acoustic":[21],"signals.We":[22],"attribute":[23],"this":[24,41],"limitation":[25],"to":[26,59,92],"modality":[27,35,75,94],"collapse,":[28],"where":[29],"models":[30],"over-rely":[31],"on":[32,114,124],"a":[33,46,151],"dominant":[34],"while":[36,131],"neglecting":[37],"complementary":[38],"cues.To":[39],"address":[40],"issue,":[42],"we":[43,77,110],"introduce":[44],"OmniCoT,":[45],"data":[47],"paradigm":[48],"that":[49,119],"interleaves":[50],"guided":[51],"tokens":[52],"(e.g.,":[53],"[vision],":[54],"[audio])":[55],"into":[56],"reasoning":[57,67,129,149],"traces":[58],"enforce":[60],"structured":[61],"evidence":[62],"extraction.To":[63],"further":[64],"internalize":[65],"the":[66,115,133,137,142],"behaviors":[68],"instilled":[69],"by":[70,108],"OmniCoT":[71,106],"facilitate":[73],"adaptive":[74],"prioritization,":[76],"propose":[78],"Dynamic":[79],"Modality-Entropy":[80],"GRPO":[81],"(DyME-GRPO),":[82],"utilizes":[84],"entropy-based":[85],"uncertainty":[86],"estimates":[87],"over":[88],"Guided":[89],"Tokens":[90],"(GTs)":[91],"regulate":[93],"usage,":[95],"thereby":[96],"mitigating":[97],"collapse":[98],"informational":[100],"redundancy.By":[101],"applying":[102],"supervised":[103],"fine-tuning":[104],"followed":[107],"DyME-GRPO,":[109],"develop":[111],"EmoOmni":[112,120],"based":[113],"Qwen2.5-Omni-7Bbackbone.Extensive":[116],"experiments":[117],"demonstrate":[118],"achieves":[121],"state-of-the-art":[122],"performance":[123],"multiple":[125],"emotion":[126],"recognition":[127],"benchmarks":[130],"preserving":[132],"general":[134],"capabilities":[135],"of":[136,144,154],"base":[138],"model.These":[139],"findings":[140],"highlight":[141],"potential":[143],"our":[145],"work":[146],"for":[147],"omni-modal":[148],"across":[150],"broader":[152],"range":[153],"tasks.":[156],"1":[157]},"counts_by_year":[],"updated_date":"2026-07-29T14:22:42.915294","created_date":"2026-07-02T00:00:00"}