{"id":"https://openalex.org/W7220896550","doi":"https://doi.org/10.48550/arxiv.2610.09757","title":"EntroPrefill: Renyi-Guided Context Pruning with Conditional Stability Guarantees for Retrieval-Augmented Generation","display_name":"EntroPrefill: Renyi-Guided Context Pruning with Conditional Stability Guarantees for Retrieval-Augmented Generation","publication_year":2026,"publication_date":"2026-10-07","ids":{"openalex":"https://openalex.org/W7220896550","doi":"https://doi.org/10.48550/arxiv.2610.09757"},"language":"en","primary_location":{"id":"pmh:oai:arXiv.org:2610.09757","doi":null,"is_oa":true,"landing_page_url":"https://arxiv.org/abs/2610.09757","pdf_url":"https://arxiv.org/pdf/2610.09757","source":{"id":"https://openalex.org/S4306400194","display_name":"arXiv (Cornell University)","issn_l":"2331-8422","issn":["2331-8422"],"is_oa":true,"is_in_doaj":false,"is_core":false,"listed_in":[],"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":null,"license_id":null,"version":"submittedVersion","is_accepted":false,"is_published":false,"raw_source_name":null,"raw_type":"text"},"volume":null,"type":"preprint","is_traditional":true,"indexed_in":["arxiv","datacite"],"open_access":{"is_oa":true,"oa_status":"green","oa_url":"https://arxiv.org/pdf/2610.09757","any_repository_has_fulltext":true},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5157605572","display_name":"Inbasekaran S","orcid":null,"observed_orcids":[]},"institutions":[],"countries":[],"is_corresponding":true,"raw_author_name":"S, Inbasekaran","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]}],"institutions":[],"countries_distinct_count":0,"institutions_distinct_count":0,"corresponding_author_ids":["https://openalex.org/A5157605572"],"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":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":0.6967999935150146,"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":0.6967999935150146,"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/T10036","display_name":"Advanced Neural Network Applications","score":0.16750000417232513,"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/T10181","display_name":"Natural Language Processing Techniques","score":0.065700002014637,"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/retrieval-augmented-generation","display_name":"retrieval-augmented generation","score":0.9409999847412109},{"id":"https://openalex.org/keywords/renyi-entropy","display_name":"R\u00e9nyi entropy","score":0.753000020980835},{"id":"https://openalex.org/keywords/grouped-query-attention","display_name":"grouped-query attention","score":0.7200000286102295},{"id":"https://openalex.org/keywords/lipschitz-constant","display_name":"Lipschitz constant","score":0.9049999713897705},{"id":"https://openalex.org/keywords/kv-cache","display_name":"KV cache","score":0.5099999904632568}],"concepts":[{"id":"https://openalex.org/C162838799","wikidata":"https://www.wikidata.org/wiki/Q596077","display_name":"Counterexample","level":2,"score":0.6929000020027161},{"id":"https://openalex.org/C22324862","wikidata":"https://www.wikidata.org/wiki/Q652707","display_name":"Lipschitz continuity","level":2,"score":0.6080999970436096},{"id":"https://openalex.org/C77553402","wikidata":"https://www.wikidata.org/wiki/Q13222579","display_name":"Upper and lower bounds","level":2,"score":0.6000000238418579},{"id":"https://openalex.org/C33923547","wikidata":"https://www.wikidata.org/wiki/Q395","display_name":"Mathematics","level":0,"score":0.5723000168800354},{"id":"https://openalex.org/C70437156","wikidata":"https://www.wikidata.org/wiki/Q7228652","display_name":"Pooling","level":2,"score":0.4650000035762787},{"id":"https://openalex.org/C95611797","wikidata":"https://www.wikidata.org/wiki/Q17502105","display_name":"Infimum and supremum","level":2,"score":0.41940000653266907},{"id":"https://openalex.org/C108010975","wikidata":"https://www.wikidata.org/wiki/Q500094","display_name":"Pruning","level":2,"score":0.41760000586509705},{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.4156999886035919},{"id":"https://openalex.org/C11413529","wikidata":"https://www.wikidata.org/wiki/Q8366","display_name":"Algorithm","level":1,"score":0.37070000171661377},{"id":"https://openalex.org/C126255220","wikidata":"https://www.wikidata.org/wiki/Q141495","display_name":"Mathematical optimization","level":1,"score":0.3580000102519989},{"id":"https://openalex.org/C2779343474","wikidata":"https://www.wikidata.org/wiki/Q3109175","display_name":"Context (archaeology)","level":2,"score":0.3431999981403351},{"id":"https://openalex.org/C78023250","wikidata":"https://www.wikidata.org/wiki/Q657596","display_name":"Unary operation","level":2,"score":0.3269999921321869},{"id":"https://openalex.org/C2780069185","wikidata":"https://www.wikidata.org/wiki/Q7977945","display_name":"Equivalence (formal languages)","level":2,"score":0.3264999985694885},{"id":"https://openalex.org/C66322947","wikidata":"https://www.wikidata.org/wiki/Q11658","display_name":"Transformer","level":3,"score":0.3249000012874603},{"id":"https://openalex.org/C2778112365","wikidata":"https://www.wikidata.org/wiki/Q3511065","display_name":"Sequence (biology)","level":2,"score":0.31929999589920044},{"id":"https://openalex.org/C2780186347","wikidata":"https://www.wikidata.org/wiki/Q11414","display_name":"Subnetwork","level":2,"score":0.3140000104904175},{"id":"https://openalex.org/C2780704645","wikidata":"https://www.wikidata.org/wiki/Q9251458","display_name":"Observer (physics)","level":2,"score":0.3133000135421753},{"id":"https://openalex.org/C112972136","wikidata":"https://www.wikidata.org/wiki/Q7595718","display_name":"Stability (learning theory)","level":2,"score":0.2996000051498413},{"id":"https://openalex.org/C167729594","wikidata":"https://www.wikidata.org/wiki/Q17736","display_name":"Axiom","level":2,"score":0.29789999127388},{"id":"https://openalex.org/C80444323","wikidata":"https://www.wikidata.org/wiki/Q2878974","display_name":"Theoretical computer science","level":1,"score":0.2865999937057495},{"id":"https://openalex.org/C48145219","wikidata":"https://www.wikidata.org/wiki/Q1335365","display_name":"Security token","level":2,"score":0.2809999883174896},{"id":"https://openalex.org/C107673813","wikidata":"https://www.wikidata.org/wiki/Q812534","display_name":"Bayesian probability","level":2,"score":0.25760000944137573},{"id":"https://openalex.org/C177918212","wikidata":"https://www.wikidata.org/wiki/Q803623","display_name":"Perturbation (astronomy)","level":2,"score":0.251800000667572}],"mesh":[],"locations_count":2,"locations":[{"id":"pmh:oai:arXiv.org:2610.09757","doi":null,"is_oa":true,"landing_page_url":"https://arxiv.org/abs/2610.09757","pdf_url":"https://arxiv.org/pdf/2610.09757","source":{"id":"https://openalex.org/S4306400194","display_name":"arXiv (Cornell University)","issn_l":"2331-8422","issn":["2331-8422"],"is_oa":true,"is_in_doaj":false,"is_core":false,"listed_in":[],"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":null,"license_id":null,"version":"submittedVersion","is_accepted":false,"is_published":false,"raw_source_name":null,"raw_type":"text"},{"id":"doi:10.48550/arxiv.2610.09757","doi":"https://doi.org/10.48550/arxiv.2610.09757","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2610.09757","pdf_url":null,"source":{"id":"https://openalex.org/S4306400194","display_name":"arXiv (Cornell University)","issn_l":"2331-8422","issn":["2331-8422"],"is_oa":true,"is_in_doaj":false,"is_core":false,"listed_in":[],"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"}],"best_oa_location":{"id":"pmh:oai:arXiv.org:2610.09757","doi":null,"is_oa":true,"landing_page_url":"https://arxiv.org/abs/2610.09757","pdf_url":"https://arxiv.org/pdf/2610.09757","source":{"id":"https://openalex.org/S4306400194","display_name":"arXiv (Cornell University)","issn_l":"2331-8422","issn":["2331-8422"],"is_oa":true,"is_in_doaj":false,"is_core":false,"listed_in":[],"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":null,"license_id":null,"version":"submittedVersion","is_accepted":false,"is_published":false,"raw_source_name":null,"raw_type":"text"},"sustainable_development_goals":[],"study_designs":[],"awards":[],"funders":[],"has_content":{"grobid_xml":true,"pdf":true},"content_urls":{"pdf":"https://content.openalex.org/works/W7220896550.pdf","grobid_xml":"https://content.openalex.org/works/W7220896550.grobid-xml"},"referenced_works_count":16,"referenced_works":["https://openalex.org/W4233413206","https://openalex.org/W4384648639","https://openalex.org/W4387294617","https://openalex.org/W4387321091","https://openalex.org/W4389518760","https://openalex.org/W4389519226","https://openalex.org/W4391045969","https://openalex.org/W4392489911","https://openalex.org/W4395482887","https://openalex.org/W4402857639","https://openalex.org/W4406071672","https://openalex.org/W4415798413","https://openalex.org/W4417280583","https://openalex.org/W7133224090","https://openalex.org/W7138092506","https://openalex.org/W7166825933"],"related_works":["https://openalex.org/W3180037928","https://openalex.org/W7166860092","https://openalex.org/W7160180282","https://openalex.org/W4399759157","https://openalex.org/W7133191579","https://openalex.org/W4406859087","https://openalex.org/W4406178928","https://openalex.org/W4382322807","https://openalex.org/W7196965237","https://openalex.org/W7204100827"],"abstract_inverted_index":{"Mid-prefill":[0],"pruning":[1,80],"can":[2],"reduce":[3],"the":[4,79,144,168],"sequence":[5],"processed":[6],"by":[7],"deeper":[8],"transformer":[9,90],"layers,":[10],"but":[11,152],"attention":[12,37,45],"concentration":[13],"alone":[14,109],"does":[15,153],"not":[16,154],"certify":[17],"that":[18,75],"discarded":[19,36],"context":[20],"is":[21,82,138],"dispensable.":[22],"We":[23,56,85],"formulate":[24],"EntroPrefill":[25],"as":[26],"a":[27,48,58,62,71,88,99],"Renyi-guided":[28],"proposal":[29],"mechanism":[30],"coupled":[31],"to":[32],"explicit":[33,94],"constraints":[34],"on":[35,66],"mass.":[38],"Sink-isolated,":[39],"regularized":[40],"head":[41],"pooling":[42],"respects":[43],"grouped-query":[44],"while":[46],"exposing":[47],"quantitative":[49],"trade-off":[50],"between":[51],"specialization":[52],"and":[53,70,98,125,128,148,172],"worst-head":[54],"coverage.":[55],"derive":[57],"mixture-to-head":[59],"deletion":[60],"envelope,":[61],"computable":[63],"upper":[64],"bound":[65,92],"feasible":[67],"token":[68],"removal,":[69],"finite-sample":[72],"observer":[73],"guarantee":[74],"remains":[76],"valid":[77],"when":[78],"layer":[81],"selected":[83],"adaptively.":[84],"then":[86],"establish":[87],"conditional":[89],"perturbation":[91],"with":[93],"sufficient":[95],"Lipschitz":[96],"constants":[97],"first-token":[100],"decision-margin":[101],"corollary.":[102],"A":[103],"counterexample":[104],"shows":[105],"why":[106],"shallow":[107],"observations":[108],"cannot":[110],"imply":[111],"an":[112,130],"unconditional":[113],"future-output":[114],"guarantee.":[115],"The":[116],"systems":[117],"analysis":[118],"distinguishes":[119],"query-head":[120],"unions,":[121],"physical":[122],"page":[123],"allocation,":[124],"KV-transfer":[126],"payload,":[127],"gives":[129],"arithmetic":[131],"break-even":[132],"condition":[133],"for":[134,164],"pruning.":[135],"This":[136],"manuscript":[137],"theoretical":[139],"in":[140],"scope:":[141],"it":[142],"defines":[143],"procedure,":[145],"its":[146,149],"assumptions,":[147,169],"formal":[150],"limits,":[151],"report":[155],"measured":[156],"acceleration":[157],"or":[158],"task-accuracy":[159],"preservation.":[160],"Experiments":[161],"are":[162],"reserved":[163],"subsequent":[165],"validation":[166],"of":[167],"approximation":[170],"tightness,":[171],"end-to-end":[173],"resource":[174],"trade-offs.":[175]},"counts_by_year":[],"updated_date":"2026-11-07T11:25:07.637180","created_date":"2026-10-09T00:00:00"}