{"id":"https://openalex.org/W7161569372","doi":"https://doi.org/10.48550/arxiv.2605.15315","title":"Context Pruning for Coding Agents via Multi-Rubric Latent Reasoning","display_name":"Context Pruning for Coding Agents via Multi-Rubric Latent Reasoning","publication_year":2026,"publication_date":"2026-05-14","ids":{"openalex":"https://openalex.org/W7161569372","doi":"https://doi.org/10.48550/arxiv.2605.15315"},"language":null,"primary_location":{"id":"doi:10.48550/arxiv.2605.15315","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2605.15315","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.2605.15315","any_repository_has_fulltext":true},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5136435668","display_name":"Jingjing Wang","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Wang, Jingjing","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5136372212","display_name":"Xiwen Chen","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Chen, Xiwen","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5136402707","display_name":"Wenhui Zhu","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Zhu, Wenhui","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5136418747","display_name":"Huayu Li","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Li, Huayu","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5136369257","display_name":"Zhengxiao He","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"He, Zhengxiao","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5136386940","display_name":"Feiyang Cai","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Cai, Feiyang","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5063347463","display_name":"Ana S. Carreon-Rascon","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Carreon-Rascon, Ana S.","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5136386901","display_name":"Xuanzhao Dong","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Dong, Xuanzhao","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"last","author":{"id":"https://openalex.org/A5136439044","display_name":"Feng Luo","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Luo, Feng","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":0.24320000410079956,"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":0.24320000410079956,"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/T12072","display_name":"Machine Learning and Algorithms","score":0.12300000339746475,"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/T10260","display_name":"Software Engineering Research","score":0.05999999865889549,"subfield":{"id":"https://openalex.org/subfields/1710","display_name":"Information Systems"},"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/relevance","display_name":"Relevance (law)","score":0.5034000277519226},{"id":"https://openalex.org/keywords/coding","display_name":"Coding (social sciences)","score":0.49939998984336853},{"id":"https://openalex.org/keywords/pruning","display_name":"Pruning","score":0.48249998688697815},{"id":"https://openalex.org/keywords/context","display_name":"Context (archaeology)","score":0.4675000011920929},{"id":"https://openalex.org/keywords/security-token","display_name":"Security token","score":0.41990000009536743},{"id":"https://openalex.org/keywords/dependency","display_name":"Dependency (UML)","score":0.41929998993873596},{"id":"https://openalex.org/keywords/language-model","display_name":"Language model","score":0.41690000891685486},{"id":"https://openalex.org/keywords/code","display_name":"Code (set theory)","score":0.40639999508857727},{"id":"https://openalex.org/keywords/encoder","display_name":"Encoder","score":0.4004000127315521},{"id":"https://openalex.org/keywords/hash-function","display_name":"Hash function","score":0.39489999413490295}],"concepts":[{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.7296000123023987},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.5857999920845032},{"id":"https://openalex.org/C158154518","wikidata":"https://www.wikidata.org/wiki/Q7310970","display_name":"Relevance (law)","level":2,"score":0.5034000277519226},{"id":"https://openalex.org/C179518139","wikidata":"https://www.wikidata.org/wiki/Q5140297","display_name":"Coding (social sciences)","level":2,"score":0.49939998984336853},{"id":"https://openalex.org/C108010975","wikidata":"https://www.wikidata.org/wiki/Q500094","display_name":"Pruning","level":2,"score":0.48249998688697815},{"id":"https://openalex.org/C2779343474","wikidata":"https://www.wikidata.org/wiki/Q3109175","display_name":"Context (archaeology)","level":2,"score":0.4675000011920929},{"id":"https://openalex.org/C119857082","wikidata":"https://www.wikidata.org/wiki/Q2539","display_name":"Machine learning","level":1,"score":0.43709999322891235},{"id":"https://openalex.org/C48145219","wikidata":"https://www.wikidata.org/wiki/Q1335365","display_name":"Security token","level":2,"score":0.41990000009536743},{"id":"https://openalex.org/C19768560","wikidata":"https://www.wikidata.org/wiki/Q320727","display_name":"Dependency (UML)","level":2,"score":0.41929998993873596},{"id":"https://openalex.org/C137293760","wikidata":"https://www.wikidata.org/wiki/Q3621696","display_name":"Language model","level":2,"score":0.41690000891685486},{"id":"https://openalex.org/C204321447","wikidata":"https://www.wikidata.org/wiki/Q30642","display_name":"Natural language processing","level":1,"score":0.41359999775886536},{"id":"https://openalex.org/C2776760102","wikidata":"https://www.wikidata.org/wiki/Q5139990","display_name":"Code (set theory)","level":3,"score":0.40639999508857727},{"id":"https://openalex.org/C118505674","wikidata":"https://www.wikidata.org/wiki/Q42586063","display_name":"Encoder","level":2,"score":0.4004000127315521},{"id":"https://openalex.org/C99138194","wikidata":"https://www.wikidata.org/wiki/Q183427","display_name":"Hash function","level":2,"score":0.39489999413490295},{"id":"https://openalex.org/C2780451532","wikidata":"https://www.wikidata.org/wiki/Q759676","display_name":"Task (project management)","level":2,"score":0.39320001006126404},{"id":"https://openalex.org/C43126263","wikidata":"https://www.wikidata.org/wiki/Q128751","display_name":"Source code","level":2,"score":0.3732999861240387},{"id":"https://openalex.org/C4679612","wikidata":"https://www.wikidata.org/wiki/Q866298","display_name":"Aggregate (composite)","level":2,"score":0.3458999991416931},{"id":"https://openalex.org/C184337299","wikidata":"https://www.wikidata.org/wiki/Q1437428","display_name":"Semantics (computer science)","level":2,"score":0.3456999957561493},{"id":"https://openalex.org/C125411270","wikidata":"https://www.wikidata.org/wiki/Q18653","display_name":"Encoding (memory)","level":2,"score":0.34470000863075256},{"id":"https://openalex.org/C63435697","wikidata":"https://www.wikidata.org/wiki/Q864135","display_name":"Binary code","level":3,"score":0.3379000127315521},{"id":"https://openalex.org/C153083717","wikidata":"https://www.wikidata.org/wiki/Q6535263","display_name":"Leverage (statistics)","level":2,"score":0.33730000257492065},{"id":"https://openalex.org/C174348530","wikidata":"https://www.wikidata.org/wiki/Q188635","display_name":"Bridging (networking)","level":2,"score":0.3181999921798706},{"id":"https://openalex.org/C90805587","wikidata":"https://www.wikidata.org/wiki/Q10944557","display_name":"Word (group theory)","level":2,"score":0.305400013923645},{"id":"https://openalex.org/C28855332","wikidata":"https://www.wikidata.org/wiki/Q198099","display_name":"Quantization (signal processing)","level":2,"score":0.30480000376701355},{"id":"https://openalex.org/C33676613","wikidata":"https://www.wikidata.org/wiki/Q13415176","display_name":"Dimension (graph theory)","level":2,"score":0.302700012922287},{"id":"https://openalex.org/C2779458634","wikidata":"https://www.wikidata.org/wiki/Q24963715","display_name":"Debiasing","level":2,"score":0.2957000136375427},{"id":"https://openalex.org/C124101348","wikidata":"https://www.wikidata.org/wiki/Q172491","display_name":"Data mining","level":1,"score":0.29159998893737793},{"id":"https://openalex.org/C188441871","wikidata":"https://www.wikidata.org/wiki/Q7554146","display_name":"Softmax function","level":3,"score":0.2912999987602234},{"id":"https://openalex.org/C206134035","wikidata":"https://www.wikidata.org/wiki/Q811525","display_name":"Treebank","level":3,"score":0.2892000079154968},{"id":"https://openalex.org/C2776214188","wikidata":"https://www.wikidata.org/wiki/Q408386","display_name":"Inference","level":2,"score":0.2847999930381775},{"id":"https://openalex.org/C104597421","wikidata":"https://www.wikidata.org/wiki/Q5148529","display_name":"Color-coding","level":2,"score":0.2838999927043915},{"id":"https://openalex.org/C127705205","wikidata":"https://www.wikidata.org/wiki/Q5748245","display_name":"Heuristics","level":2,"score":0.28130000829696655},{"id":"https://openalex.org/C48372109","wikidata":"https://www.wikidata.org/wiki/Q3913","display_name":"Binary number","level":2,"score":0.27959999442100525},{"id":"https://openalex.org/C2779960059","wikidata":"https://www.wikidata.org/wiki/Q7113681","display_name":"Overhead (engineering)","level":2,"score":0.2773999869823456},{"id":"https://openalex.org/C183322885","wikidata":"https://www.wikidata.org/wiki/Q17007702","display_name":"Context model","level":3,"score":0.26980000734329224},{"id":"https://openalex.org/C118930307","wikidata":"https://www.wikidata.org/wiki/Q600590","display_name":"Tuple","level":2,"score":0.26829999685287476},{"id":"https://openalex.org/C70518039","wikidata":"https://www.wikidata.org/wiki/Q16000077","display_name":"Dimensionality reduction","level":2,"score":0.2581999897956848},{"id":"https://openalex.org/C152565575","wikidata":"https://www.wikidata.org/wiki/Q1124538","display_name":"Conditional random field","level":2,"score":0.25609999895095825},{"id":"https://openalex.org/C22367795","wikidata":"https://www.wikidata.org/wiki/Q7625208","display_name":"Structured prediction","level":2,"score":0.25589999556541443},{"id":"https://openalex.org/C57273362","wikidata":"https://www.wikidata.org/wiki/Q576722","display_name":"Decoding methods","level":2,"score":0.25440001487731934}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.48550/arxiv.2605.15315","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2605.15315","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.2605.15315","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2605.15315","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":[{"score":0.7519993185997009,"id":"https://metadata.un.org/sdg/4","display_name":"Quality Education"}],"awards":[],"funders":[],"has_content":{"grobid_xml":false,"pdf":false},"content_urls":null,"referenced_works_count":0,"referenced_works":[],"related_works":[],"abstract_inverted_index":{"LLM-powered":[0],"coding":[1],"agents":[2],"spend":[3],"the":[4,16,22,117,122,130,134,151,161,227],"majority":[5],"of":[6,15,40,193],"their":[7],"token":[8],"budget":[9],"reading":[10],"repository":[11],"files,":[12],"yet":[13],"much":[14],"retrieved":[17],"code":[18,41,89],"is":[19,222],"irrelevant":[20],"to":[21,201,215],"task":[23],"at":[24],"hand.":[25],"Existing":[26],"learned":[27],"pruners":[28],"compress":[29],"this":[30,53],"context":[31],"with":[32,107],"a":[33,56,59,83,104,125],"single-objective":[34],"sequence":[35],"labeler,":[36],"collapsing":[37],"all":[38],"facets":[39],"relevance":[42,90],"into":[43,91],"one":[44,47],"score":[45],"and":[46,73,98,124,209,229],"transition":[48,62,109],"matrix.":[49],"We":[50,78],"show":[51,188],"that":[52,87,189],"formulation":[54],"creates":[55],"modeling":[57],"bottleneck:":[58],"single":[60],"CRF":[61,106,127],"prior":[63],"must":[64],"serve":[65],"heterogeneous":[66],"retention":[67],"patterns,":[68],"including":[69],"contiguous":[70],"semantic":[71,96],"spans":[72],"sparse":[74],"structural":[75],"support":[76],"lines.":[77],"propose":[79],"LaMR":[80,170,190],"(Latent":[81],"Multi-Rubric),":[82],"structured":[84],"pruning":[85],"framework":[86],"decomposes":[88],"two":[92],"interpretable":[93],"quality":[94],"dimensions,":[95],"evidence":[97],"dependency":[99],"support,":[100],"each":[101,140],"modeled":[102],"by":[103,213,225],"dedicated":[105],"dimension-specific":[108],"dynamics.":[110],"A":[111],"mixture-of-experts":[112],"gating":[113],"network":[114],"dynamically":[115],"weights":[116],"per-rubric":[118],"emissions":[119,132],"conditioned":[120],"on":[121,129,180,205,217],"query,":[123],"final":[126],"layer":[128],"fused":[131],"produces":[133],"aggregate":[135],"keep-or-prune":[136],"decision.":[137],"To":[138],"supervise":[139],"dimension":[141],"without":[142],"additional":[143],"annotation":[144],"cost,":[145],"we":[146],"derive":[147],"multi-rubric":[148],"labels":[149],"from":[150],"existing":[152],"training":[153],"corpus":[154],"via":[155],"AST-based":[156],"program":[157],"analysis,":[158],"simultaneously":[159],"denoising":[160,226],"teacher's":[162],"binary":[163],"labels.":[164],"By":[165],"effectively":[166],"filtering":[167],"distracting":[168],"noise,":[169],"frequently":[171,223],"matches":[172],"or":[173],"even":[174],"outperforms":[175],"unpruned":[176],"full-context":[177],"baselines.":[178],"Experiments":[179],"four":[181],"benchmarks":[182],"(SWE-Bench":[183],"Verified,":[184],"SWE-QA,":[185],"LCC,":[186],"LongCodeQA)":[187],"wins":[191],"12":[192],"16":[194],"head-to-head":[195],"multi-turn":[196,206],"comparisons.":[197],"It":[198],"saves":[199],"up":[200,214],"31%":[202],"more":[203],"tokens":[204],"agent":[207],"tasks":[208],"improves":[210],"Exact":[211],"Match":[212],"+3.5":[216],"single-turn":[218],"tasks,":[219],"while":[220],"performance":[221],"enhanced":[224],"context,":[228],"any":[230],"remaining":[231],"drops":[232],"are":[233],"marginal.":[234]},"counts_by_year":[],"updated_date":"2026-07-01T06:00:48.157686","created_date":"2026-05-19T00:00:00"}