{"id":"https://openalex.org/W7133310305","doi":"https://doi.org/10.48550/arxiv.2603.01914","title":"AdaPonderLM: Gated Pondering Language Models with Token-Wise Adaptive Depth","display_name":"AdaPonderLM: Gated Pondering Language Models with Token-Wise Adaptive Depth","publication_year":2026,"publication_date":"2026-03-02","ids":{"openalex":"https://openalex.org/W7133310305","doi":"https://doi.org/10.48550/arxiv.2603.01914"},"language":null,"primary_location":{"id":"doi:10.48550/arxiv.2603.01914","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2603.01914","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":"cc-by","license_id":"https://openalex.org/licenses/cc-by","version":null,"is_accepted":false,"is_published":false,"raw_source_name":null,"raw_type":"article"},"type":"preprint","indexed_in":["datacite"],"open_access":{"is_oa":true,"oa_status":"green","oa_url":"https://doi.org/10.48550/arxiv.2603.01914","any_repository_has_fulltext":true},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5075075278","display_name":"Shixiang Song","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Song, Shixiang","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5128028780","display_name":"He Li","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Li, He","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5127976669","display_name":"Zitong Wang","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Wang, Zitong","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5125504206","display_name":"Boyi Zeng","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Zeng, Boyi","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5125566081","display_name":"Feichen Song","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Song, Feichen","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5128026815","display_name":"Yixuan Wang","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Wang, Yixuan","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5125893253","display_name":"Zhiqin John Xu","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Xu, Zhiqin John","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5127907753","display_name":"Ziwei He","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"He, Ziwei","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"last","author":{"id":null,"display_name":"Lin, Zhouhan","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Lin, Zhouhan","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]}],"institutions":[],"countries_distinct_count":0,"institutions_distinct_count":9,"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/T10028","display_name":"Topic Modeling","score":0.3675999939441681,"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/T10028","display_name":"Topic Modeling","score":0.3675999939441681,"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/T11948","display_name":"Machine Learning in Materials Science","score":0.1607999950647354,"subfield":{"id":"https://openalex.org/subfields/2505","display_name":"Materials Chemistry"},"field":{"id":"https://openalex.org/fields/25","display_name":"Materials Science"},"domain":{"id":"https://openalex.org/domains/3","display_name":"Physical Sciences"}},{"id":"https://openalex.org/T10181","display_name":"Natural Language Processing Techniques","score":0.06960000097751617,"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/perplexity","display_name":"Perplexity","score":0.7555000185966492},{"id":"https://openalex.org/keywords/language-model","display_name":"Language model","score":0.7419999837875366},{"id":"https://openalex.org/keywords/computation","display_name":"Computation","score":0.6578999757766724},{"id":"https://openalex.org/keywords/security-token","display_name":"Security token","score":0.5965999960899353},{"id":"https://openalex.org/keywords/pruning","display_name":"Pruning","score":0.4968000054359436},{"id":"https://openalex.org/keywords/inference","display_name":"Inference","score":0.489300012588501},{"id":"https://openalex.org/keywords/operand","display_name":"Operand","score":0.41100001335144043},{"id":"https://openalex.org/keywords/transformer","display_name":"Transformer","score":0.4066999852657318}],"concepts":[{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.8263999819755554},{"id":"https://openalex.org/C100279451","wikidata":"https://www.wikidata.org/wiki/Q372193","display_name":"Perplexity","level":3,"score":0.7555000185966492},{"id":"https://openalex.org/C137293760","wikidata":"https://www.wikidata.org/wiki/Q3621696","display_name":"Language model","level":2,"score":0.7419999837875366},{"id":"https://openalex.org/C45374587","wikidata":"https://www.wikidata.org/wiki/Q12525525","display_name":"Computation","level":2,"score":0.6578999757766724},{"id":"https://openalex.org/C48145219","wikidata":"https://www.wikidata.org/wiki/Q1335365","display_name":"Security token","level":2,"score":0.5965999960899353},{"id":"https://openalex.org/C108010975","wikidata":"https://www.wikidata.org/wiki/Q500094","display_name":"Pruning","level":2,"score":0.4968000054359436},{"id":"https://openalex.org/C2776214188","wikidata":"https://www.wikidata.org/wiki/Q408386","display_name":"Inference","level":2,"score":0.489300012588501},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.4875999987125397},{"id":"https://openalex.org/C55526617","wikidata":"https://www.wikidata.org/wiki/Q719375","display_name":"Operand","level":2,"score":0.41100001335144043},{"id":"https://openalex.org/C66322947","wikidata":"https://www.wikidata.org/wiki/Q11658","display_name":"Transformer","level":3,"score":0.4066999852657318},{"id":"https://openalex.org/C2776436953","wikidata":"https://www.wikidata.org/wiki/Q5163215","display_name":"Consistency (knowledge bases)","level":2,"score":0.3662000000476837},{"id":"https://openalex.org/C206588197","wikidata":"https://www.wikidata.org/wiki/Q846574","display_name":"Reuse","level":2,"score":0.35929998755455017},{"id":"https://openalex.org/C44291984","wikidata":"https://www.wikidata.org/wiki/Q1074173","display_name":"Question answering","level":2,"score":0.3549000024795532},{"id":"https://openalex.org/C115537543","wikidata":"https://www.wikidata.org/wiki/Q165596","display_name":"Cache","level":2,"score":0.3199999928474426},{"id":"https://openalex.org/C2780069185","wikidata":"https://www.wikidata.org/wiki/Q7977945","display_name":"Equivalence (formal languages)","level":2,"score":0.3149000108242035},{"id":"https://openalex.org/C37279795","wikidata":"https://www.wikidata.org/wiki/Q2492305","display_name":"Consistency model","level":3,"score":0.3019999861717224},{"id":"https://openalex.org/C128099668","wikidata":"https://www.wikidata.org/wiki/Q573952","display_name":"Lazy evaluation","level":3,"score":0.29899999499320984},{"id":"https://openalex.org/C48044578","wikidata":"https://www.wikidata.org/wiki/Q727490","display_name":"Scalability","level":2,"score":0.27469998598098755},{"id":"https://openalex.org/C99138194","wikidata":"https://www.wikidata.org/wiki/Q183427","display_name":"Hash function","level":2,"score":0.26919999718666077},{"id":"https://openalex.org/C153083717","wikidata":"https://www.wikidata.org/wiki/Q6535263","display_name":"Leverage (statistics)","level":2,"score":0.2628999948501587},{"id":"https://openalex.org/C11413529","wikidata":"https://www.wikidata.org/wiki/Q8366","display_name":"Algorithm","level":1,"score":0.25060001015663147}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.48550/arxiv.2603.01914","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2603.01914","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":"cc-by","license_id":"https://openalex.org/licenses/cc-by","version":null,"is_accepted":false,"is_published":null,"raw_source_name":null,"raw_type":"article"}],"best_oa_location":{"id":"doi:10.48550/arxiv.2603.01914","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2603.01914","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":"cc-by","license_id":"https://openalex.org/licenses/cc-by","version":null,"is_accepted":false,"is_published":false,"raw_source_name":null,"raw_type":"article"},"sustainable_development_goals":[],"awards":[],"funders":[],"has_content":{"grobid_xml":false,"pdf":false},"content_urls":null,"referenced_works_count":0,"referenced_works":[],"related_works":[],"abstract_inverted_index":{"Test-time":[0],"scaling":[1],"via":[2],"recurrent/iterative":[3],"Transformers":[4],"enables":[5],"large":[6],"language":[7,52,128],"models":[8],"to":[9,77,109,114,144,173],"spend":[10],"more":[11,142],"computation":[12,143,150],"at":[13,122],"inference,":[14],"but":[15],"most":[16],"pretrained":[17],"recurrent":[18,51],"LMs":[19],"run":[20],"a":[21,49,73,86,154],"fixed":[22,167],"number":[23],"of":[24,39],"iterations,":[25],"wasting":[26],"compute":[27,121,172],"on":[28],"easy":[29],"tokens":[30,176],"and":[31,43,84,101,112,131],"lacking":[32],"token-wise":[33,56],"adaptivity.":[34],"Following":[35],"the":[36,138,161,174],"core":[37],"idea":[38],"Adaptive":[40],"Computation":[41],"Time(ACT)":[42],"Early":[44],"Exit(EE),":[45],"we":[46],"propose":[47],"AdaPonderLM,":[48],"self-supervised":[50,156],"model":[53],"that":[54,90],"learns":[55],"early":[57],"exiting":[58],"during":[59],"pretraining":[60],"without":[61],"manually":[62],"tuned":[63],"per-token/per-layer":[64],"pruning":[65],"ratios.":[66],"AdaPonderLM":[67,118,170],"uses":[68],"iteration-specific":[69],"MLP":[70],"gates":[71,140],"with":[72],"monotonic":[74],"halting":[75,163],"mask":[76],"decide":[78],"when":[79],"each":[80],"token":[81],"stops":[82],"recurring,":[83],"introduces":[85],"KV":[87],"reuse":[88],"mechanism":[89],"reuses":[91],"cached":[92],"key/value":[93],"states":[94],"for":[95],"halted":[96],"tokens,":[97,147],"ensuring":[98],"train--test":[99],"consistency":[100],"practical":[102],"acceleration.":[103],"Across":[104],"Pythia":[105],"backbones":[106],"from":[107],"70M":[108],"410M":[110],"(pretraining)":[111],"up":[113],"2.8B":[115],"(continued":[116],"pretraining),":[117],"reduces":[119],"inference":[120],"about":[123],"10%":[124],"while":[125],"maintaining":[126],"comparable":[127],"modeling":[129],"perplexity":[130],"competitive":[132],"downstream":[133],"accuracy.":[134],"Our":[135],"analysis":[136],"shows":[137],"learned":[139,162],"allocate":[141],"high-NLL":[145],"(hard)":[146],"exhibiting":[148],"adaptive":[149],"time":[151],"behavior":[152],"in":[153],"fully":[155],"setting.":[157],"Meanwhile,":[158],"under":[159],"iso-FLOPs,":[160],"policy":[164],"consistently":[165],"outperforms":[166],"pruning,":[168],"showing":[169],"allocates":[171],"right":[175],"rather":[177],"than":[178],"just":[179],"reducing":[180],"average":[181],"depth.":[182]},"counts_by_year":[],"updated_date":"2026-06-11T09:08:48.828518","created_date":"2026-03-04T00:00:00"}