{"id":"https://openalex.org/W4417073324","doi":"https://doi.org/10.1145/3743093.3771045","title":"Gradient Shaping Beyond Clipping: A Functional Perspective on Update Magnitude Control","display_name":"Gradient Shaping Beyond Clipping: A Functional Perspective on Update Magnitude Control","publication_year":2025,"publication_date":"2025-12-06","ids":{"openalex":"https://openalex.org/W4417073324","doi":"https://doi.org/10.1145/3743093.3771045"},"language":null,"primary_location":{"id":"doi:10.1145/3743093.3771045","is_oa":true,"landing_page_url":"https://doi.org/10.1145/3743093.3771045","pdf_url":"https://dl.acm.org/doi/pdf/10.1145/3743093.3771045","source":null,"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"Proceedings of the 7th ACM International Conference on Multimedia in Asia","raw_type":"proceedings-article"},"type":"article","indexed_in":["crossref"],"open_access":{"is_oa":true,"oa_status":"gold","oa_url":"https://dl.acm.org/doi/pdf/10.1145/3743093.3771045","any_repository_has_fulltext":null},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5102631359","display_name":"Haochen You","orcid":"https://orcid.org/0009-0008-9178-2912"},"institutions":[{"id":"https://openalex.org/I174216632","display_name":"City University of New York","ror":"https://ror.org/00453a208","country_code":"US","type":"education","lineage":["https://openalex.org/I174216632"]}],"countries":["US"],"is_corresponding":false,"raw_author_name":"Haochen You","raw_affiliation_strings":["Graduate School of Arts and Sciences, Columbia University, New York City, New York, USA"],"raw_orcid":"https://orcid.org/0009-0008-9178-2912","affiliations":[{"raw_affiliation_string":"Graduate School of Arts and Sciences, Columbia University, New York City, New York, USA","institution_ids":["https://openalex.org/I174216632"]}]},{"author_position":"last","author":{"id":"https://openalex.org/A5075692649","display_name":"Baojing Liu","orcid":"https://orcid.org/0009-0007-1444-7267"},"institutions":[{"id":"https://openalex.org/I4210164386","display_name":"Hebei Science and Technology Department","ror":"https://ror.org/05k812a28","country_code":"CN","type":"government","lineage":["https://openalex.org/I4210164386"]}],"countries":["CN"],"is_corresponding":false,"raw_author_name":"Baojing Liu","raw_affiliation_strings":["School of Artificial Intelligence, Hebei Institute of Communications, Shijiazhuang, Hebei Province, China"],"raw_orcid":"https://orcid.org/0009-0007-1444-7267","affiliations":[{"raw_affiliation_string":"School of Artificial Intelligence, Hebei Institute of Communications, Shijiazhuang, Hebei Province, China","institution_ids":["https://openalex.org/I4210164386"]}]}],"institutions":[],"countries_distinct_count":2,"institutions_distinct_count":2,"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":0.18466266,"is_in_top_1_percent":false,"is_in_top_10_percent":false},"cited_by_percentile_year":null,"biblio":{"volume":null,"issue":null,"first_page":"1","last_page":"7"},"is_retracted":false,"is_paratext":false,"is_xpac":false,"primary_topic":{"id":"https://openalex.org/T11612","display_name":"Stochastic Gradient Optimization Techniques","score":0.11980000138282776,"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/T11612","display_name":"Stochastic Gradient Optimization Techniques","score":0.11980000138282776,"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/T10775","display_name":"Generative Adversarial Networks and Image Synthesis","score":0.10999999940395355,"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/T11307","display_name":"Domain Adaptation and Few-Shot Learning","score":0.09120000153779984,"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/clipping","display_name":"Clipping (morphology)","score":0.6570000052452087},{"id":"https://openalex.org/keywords/robustness","display_name":"Robustness (evolution)","score":0.6506999731063843},{"id":"https://openalex.org/keywords/perspective","display_name":"Perspective (graphical)","score":0.5945000052452087},{"id":"https://openalex.org/keywords/gradient-descent","display_name":"Gradient descent","score":0.5080999732017517},{"id":"https://openalex.org/keywords/flexibility","display_name":"Flexibility (engineering)","score":0.49149999022483826},{"id":"https://openalex.org/keywords/differentiable-function","display_name":"Differentiable function","score":0.4690000116825104},{"id":"https://openalex.org/keywords/scale","display_name":"Scale (ratio)","score":0.4016000032424927},{"id":"https://openalex.org/keywords/control-theory","display_name":"Control theory (sociology)","score":0.37220001220703125}],"concepts":[{"id":"https://openalex.org/C2776848632","wikidata":"https://www.wikidata.org/wiki/Q853463","display_name":"Clipping (morphology)","level":2,"score":0.6570000052452087},{"id":"https://openalex.org/C63479239","wikidata":"https://www.wikidata.org/wiki/Q7353546","display_name":"Robustness (evolution)","level":3,"score":0.6506999731063843},{"id":"https://openalex.org/C12713177","wikidata":"https://www.wikidata.org/wiki/Q1900281","display_name":"Perspective (graphical)","level":2,"score":0.5945000052452087},{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.5501000285148621},{"id":"https://openalex.org/C153258448","wikidata":"https://www.wikidata.org/wiki/Q1199743","display_name":"Gradient descent","level":3,"score":0.5080999732017517},{"id":"https://openalex.org/C2780598303","wikidata":"https://www.wikidata.org/wiki/Q65921492","display_name":"Flexibility (engineering)","level":2,"score":0.49149999022483826},{"id":"https://openalex.org/C202615002","wikidata":"https://www.wikidata.org/wiki/Q783507","display_name":"Differentiable function","level":2,"score":0.4690000116825104},{"id":"https://openalex.org/C2778755073","wikidata":"https://www.wikidata.org/wiki/Q10858537","display_name":"Scale (ratio)","level":2,"score":0.4016000032424927},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.3978999853134155},{"id":"https://openalex.org/C11413529","wikidata":"https://www.wikidata.org/wiki/Q8366","display_name":"Algorithm","level":1,"score":0.3921999931335449},{"id":"https://openalex.org/C47446073","wikidata":"https://www.wikidata.org/wiki/Q5165890","display_name":"Control theory (sociology)","level":3,"score":0.37220001220703125},{"id":"https://openalex.org/C33923547","wikidata":"https://www.wikidata.org/wiki/Q395","display_name":"Mathematics","level":0,"score":0.3481999933719635},{"id":"https://openalex.org/C73301696","wikidata":"https://www.wikidata.org/wiki/Q5469984","display_name":"Formalism (music)","level":3,"score":0.31290000677108765},{"id":"https://openalex.org/C115680565","wikidata":"https://www.wikidata.org/wiki/Q5977448","display_name":"Gradient method","level":2,"score":0.31290000677108765},{"id":"https://openalex.org/C17020691","wikidata":"https://www.wikidata.org/wiki/Q139677","display_name":"Operator (biology)","level":5,"score":0.30970001220703125},{"id":"https://openalex.org/C2780980858","wikidata":"https://www.wikidata.org/wiki/Q110022","display_name":"Dual (grammatical number)","level":2,"score":0.3001999855041504},{"id":"https://openalex.org/C28719098","wikidata":"https://www.wikidata.org/wiki/Q44946","display_name":"Point (geometry)","level":2,"score":0.2980000078678131},{"id":"https://openalex.org/C115961682","wikidata":"https://www.wikidata.org/wiki/Q860623","display_name":"Image (mathematics)","level":2,"score":0.28769999742507935},{"id":"https://openalex.org/C107464732","wikidata":"https://www.wikidata.org/wiki/Q235781","display_name":"Adaptive control","level":3,"score":0.27549999952316284},{"id":"https://openalex.org/C96133863","wikidata":"https://www.wikidata.org/wiki/Q6913458","display_name":"Morphological gradient","level":5,"score":0.274399995803833},{"id":"https://openalex.org/C110121322","wikidata":"https://www.wikidata.org/wiki/Q865811","display_name":"Distribution (mathematics)","level":2,"score":0.271699994802475},{"id":"https://openalex.org/C2777303404","wikidata":"https://www.wikidata.org/wiki/Q759757","display_name":"Convergence (economics)","level":2,"score":0.25200000405311584}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.1145/3743093.3771045","is_oa":true,"landing_page_url":"https://doi.org/10.1145/3743093.3771045","pdf_url":"https://dl.acm.org/doi/pdf/10.1145/3743093.3771045","source":null,"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"Proceedings of the 7th ACM International Conference on Multimedia in Asia","raw_type":"proceedings-article"}],"best_oa_location":{"id":"doi:10.1145/3743093.3771045","is_oa":true,"landing_page_url":"https://doi.org/10.1145/3743093.3771045","pdf_url":"https://dl.acm.org/doi/pdf/10.1145/3743093.3771045","source":null,"license":null,"license_id":null,"version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"Proceedings of the 7th ACM International Conference on Multimedia in Asia","raw_type":"proceedings-article"},"sustainable_development_goals":[],"awards":[],"funders":[],"has_content":{"pdf":true,"grobid_xml":true},"content_urls":{"pdf":"https://content.openalex.org/works/W4417073324.pdf","grobid_xml":"https://content.openalex.org/works/W4417073324.grobid-xml"},"referenced_works_count":6,"referenced_works":["https://openalex.org/W1994616650","https://openalex.org/W2251939518","https://openalex.org/W4401863489","https://openalex.org/W4410237074","https://openalex.org/W4411631879","https://openalex.org/W4415799145"],"related_works":[],"abstract_inverted_index":{"Gradient":[0],"clipping":[1,39,68],"is":[2],"widely":[3],"used":[4],"to":[5,57,85],"stabilize":[6],"deep":[7],"network":[8],"training,":[9],"but":[10],"its":[11],"formulation":[12],"as":[13,71],"a":[14,34,62,82],"hard,":[15],"fixed":[16],"threshold":[17],"limits":[18],"flexibility":[19],"and":[20,32,53,69,92,101],"ignores":[21],"gradient":[22,43,48],"distribution":[23],"dynamics.":[24],"We":[25],"propose":[26],"SPAMP":[27,45,97],"(Statistical":[28],"Per-layer":[29],"Adaptive":[30],"Modulation":[31],"Projection),":[33],"unified":[35],"framework":[36],"that":[37,96],"generalizes":[38],"into":[40],"smooth,":[41],"per-layer":[42],"shaping.":[44],"tracks":[46],"local":[47],"statistics,":[49],"dynamically":[50],"estimates":[51],"thresholds,":[52],"applies":[54],"power-based":[55],"transformations":[56],"modulate":[58],"update":[59,78],"magnitudes":[60],"in":[61],"differentiable":[63],"manner.":[64],"This":[65],"perspective":[66],"recasts":[67],"warmup":[70],"dual":[72],"mechanisms":[73],"for":[74],"controlling":[75],"the":[76],"effective":[77],"scale":[79],"\u03b7t\u2016gt\u2016,":[80],"offering":[81],"principled":[83],"alternative":[84],"rigid":[86],"heuristics.":[87],"Extensive":[88],"experiments":[89],"across":[90],"image":[91],"language":[93],"tasks":[94],"demonstrate":[95],"improves":[98],"stability,":[99],"convergence,":[100],"robustness":[102],"over":[103],"existing":[104],"methods.":[105]},"counts_by_year":[],"updated_date":"2026-06-11T09:08:48.828518","created_date":"2025-12-06T00:00:00"}