{"id":"https://openalex.org/W7163887246","doi":"https://doi.org/10.48550/arxiv.2606.06574","title":"Skip a Layer or Loop It? Learning Program-of-Layers in LLMs","display_name":"Skip a Layer or Loop It? Learning Program-of-Layers in LLMs","publication_year":2026,"publication_date":"2026-06-04","ids":{"openalex":"https://openalex.org/W7163887246","doi":"https://doi.org/10.48550/arxiv.2606.06574"},"language":null,"primary_location":{"id":"doi:10.48550/arxiv.2606.06574","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2606.06574","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":"Preprint"},"type":"preprint","indexed_in":["datacite"],"open_access":{"is_oa":true,"oa_status":"green","oa_url":"https://doi.org/10.48550/arxiv.2606.06574","any_repository_has_fulltext":true},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5138150675","display_name":"Ziyue Li","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Li, Ziyue","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5138192674","display_name":"Yang Li","orcid":"https://orcid.org/0000-0002-5066-1734"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Li, Yang","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"last","author":{"id":"https://openalex.org/A5138159671","display_name":"Tianyi Zhou","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Zhou, Tianyi","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":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":0.19480000436306,"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":0.19480000436306,"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":0.10159999877214432,"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/T11714","display_name":"Multimodal Machine Learning Applications","score":0.09019999951124191,"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/inference","display_name":"Inference","score":0.8251000046730042},{"id":"https://openalex.org/keywords/layer","display_name":"Layer (electronics)","score":0.6294999718666077},{"id":"https://openalex.org/keywords/computation","display_name":"Computation","score":0.6021999716758728},{"id":"https://openalex.org/keywords/loop","display_name":"Loop (graph theory)","score":0.46160000562667847},{"id":"https://openalex.org/keywords/polar","display_name":"Polar","score":0.4065999984741211},{"id":"https://openalex.org/keywords/execution-time","display_name":"Execution time","score":0.287200003862381}],"concepts":[{"id":"https://openalex.org/C2776214188","wikidata":"https://www.wikidata.org/wiki/Q408386","display_name":"Inference","level":2,"score":0.8251000046730042},{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.7350000143051147},{"id":"https://openalex.org/C2779227376","wikidata":"https://www.wikidata.org/wiki/Q6505497","display_name":"Layer (electronics)","level":2,"score":0.6294999718666077},{"id":"https://openalex.org/C45374587","wikidata":"https://www.wikidata.org/wiki/Q12525525","display_name":"Computation","level":2,"score":0.6021999716758728},{"id":"https://openalex.org/C184670325","wikidata":"https://www.wikidata.org/wiki/Q512604","display_name":"Loop (graph theory)","level":2,"score":0.46160000562667847},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.44029998779296875},{"id":"https://openalex.org/C29705727","wikidata":"https://www.wikidata.org/wiki/Q294562","display_name":"Polar","level":2,"score":0.4065999984741211},{"id":"https://openalex.org/C119857082","wikidata":"https://www.wikidata.org/wiki/Q2539","display_name":"Machine learning","level":1,"score":0.39989998936653137},{"id":"https://openalex.org/C11413529","wikidata":"https://www.wikidata.org/wiki/Q8366","display_name":"Algorithm","level":1,"score":0.3573000133037567},{"id":"https://openalex.org/C199360897","wikidata":"https://www.wikidata.org/wiki/Q9143","display_name":"Programming language","level":1,"score":0.3237999975681305},{"id":"https://openalex.org/C80444323","wikidata":"https://www.wikidata.org/wiki/Q2878974","display_name":"Theoretical computer science","level":1,"score":0.3028999865055084},{"id":"https://openalex.org/C2989134064","wikidata":"https://www.wikidata.org/wiki/Q288510","display_name":"Execution time","level":2,"score":0.287200003862381},{"id":"https://openalex.org/C51632099","wikidata":"https://www.wikidata.org/wiki/Q3985153","display_name":"Training set","level":2,"score":0.2619999945163727},{"id":"https://openalex.org/C3017489831","wikidata":"https://www.wikidata.org/wiki/Q2393193","display_name":"Running time","level":2,"score":0.26100000739097595},{"id":"https://openalex.org/C186886427","wikidata":"https://www.wikidata.org/wiki/Q5441213","display_name":"Feedback loop","level":2,"score":0.25839999318122864},{"id":"https://openalex.org/C3746660","wikidata":"https://www.wikidata.org/wiki/Q1068763","display_name":"Rule of inference","level":2,"score":0.250900000333786}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.48550/arxiv.2606.06574","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2606.06574","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":"Preprint"}],"best_oa_location":{"id":"doi:10.48550/arxiv.2606.06574","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2606.06574","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":"Preprint"},"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":{"Large":[0],"language":[1],"models":[2],"(LLMs)":[3],"perform":[4],"inference":[5,84,137],"by":[6,74],"following":[7],"a":[8,44,103,163],"fixed":[9],"depth":[10],"and":[11,37,138,147],"order,":[12],"non-recurrent":[13],"execution":[14,112,160],"of":[15,23,67,166],"all":[16],"layers.":[17,79],"We":[18],"reveal":[19],"the":[20,59,68,91],"wide":[21],"existence":[22],"training-free,":[24],"flexible,":[25],"dynamic":[26],"program-of-layers":[27],"(PoLar),":[28],"where":[29],"pretrained":[30,119],"layers":[31,120],"can":[32,57,71],"be":[33,72],"packed":[34],"as":[35],"modules":[36],"then":[38],"skipped":[39],"or":[40,61,117],"looped":[41],"to":[42,110],"form":[43],"customized":[45],"program":[46,55],"for":[47,121],"each":[48,122],"input.":[49,123],"For":[50],"most":[51],"inputs,":[52],"substantially":[53],"shorter":[54],"executions":[56],"achieve":[58,97],"same":[60],"better":[62],"accuracy,":[63],"while":[64,143],"incorrect":[65],"predictions":[66],"original":[69],"LLM":[70],"corrected":[73],"alternative":[75],"programs":[76,113],"with":[77],"fewer":[78,145],"These":[80],"observations":[81],"indicate":[82],"that":[83,114,130,148,158],"admits":[85],"multiple":[86],"valid":[87],"latent":[88,169],"computations":[89],"beyond":[90],"standard":[92,136],"forward":[93],"pass.":[94],"To":[95],"efficiently":[96],"PoLar":[98,105,131],"in":[99],"practice,":[100],"we":[101],"propose":[102],"lightweight":[104],"prediction":[106],"network,":[107],"which":[108],"learns":[109],"generate":[111],"dynamically":[115],"skip":[116],"repeat":[118],"Experiments":[124],"on":[125],"mathematical":[126],"reasoning":[127,170],"benchmarks":[128],"demonstrate":[129],"consistently":[132],"improves":[133],"accuracy":[134],"over":[135],"prior":[139],"dynamic-depth":[140],"methods,":[141],"often":[142],"executing":[144],"layers,":[146],"these":[149],"gains":[150],"persist":[151],"under":[152],"out-of-distribution":[153],"evaluation.":[154],"Our":[155],"results":[156],"suggest":[157],"fixed-depth":[159],"captures":[161],"only":[162],"narrow":[164],"subset":[165],"an":[167],"LLM's":[168],"capacity.":[171]},"counts_by_year":[],"updated_date":"2026-07-01T06:00:48.157686","created_date":"2026-06-09T00:00:00"}