{"id":"https://openalex.org/W7163182634","doi":"https://doi.org/10.48550/arxiv.2606.01923","title":"Resonant Context Anchoring: Decoupling Attention Routing and Signal Gain at Inference Time","display_name":"Resonant Context Anchoring: Decoupling Attention Routing and Signal Gain at Inference Time","publication_year":2026,"publication_date":"2026-06-01","ids":{"openalex":"https://openalex.org/W7163182634","doi":"https://doi.org/10.48550/arxiv.2606.01923"},"language":null,"primary_location":{"id":"doi:10.48550/arxiv.2606.01923","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2606.01923","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":null,"license_id":null,"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.2606.01923","any_repository_has_fulltext":true},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5137664386","display_name":"Mingkuan Zhao","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Zhao, Mingkuan","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5137661174","display_name":"Yide Gao","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Gao, Yide","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5137686994","display_name":"Wentao Hu","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Hu, Wentao","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5137693638","display_name":"Suquan Chen","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Chen, Suquan","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5068378237","display_name":"Tianqi Huang","orcid":"https://orcid.org/0009-0008-4028-8244"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Huang, Tianchen","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5137678057","display_name":"Zhenhua An","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"An, Zhenhua","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5137684170","display_name":"Zetao Chang","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Chang, Zetao","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5137636657","display_name":"Xiayu Sun","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Sun, Xiayu","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"last","author":{"id":"https://openalex.org/A5137641739","display_name":"Yuheng Min","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Min, Yuheng","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":0.2540000081062317,"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":0.2540000081062317,"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":0.13269999623298645,"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":0.11990000307559967,"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/language-model","display_name":"Language model","score":0.5239999890327454},{"id":"https://openalex.org/keywords/inference","display_name":"Inference","score":0.47929999232292175},{"id":"https://openalex.org/keywords/decoding-methods","display_name":"Decoding methods","score":0.46140000224113464},{"id":"https://openalex.org/keywords/context","display_name":"Context (archaeology)","score":0.4174000024795532},{"id":"https://openalex.org/keywords/parametric-statistics","display_name":"Parametric statistics","score":0.39640000462532043},{"id":"https://openalex.org/keywords/unavailability","display_name":"Unavailability","score":0.3479999899864197},{"id":"https://openalex.org/keywords/residual","display_name":"Residual","score":0.3450999855995178},{"id":"https://openalex.org/keywords/robustness","display_name":"Robustness (evolution)","score":0.3440000116825104},{"id":"https://openalex.org/keywords/metric","display_name":"Metric (unit)","score":0.30239999294281006}],"concepts":[{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.7063999772071838},{"id":"https://openalex.org/C137293760","wikidata":"https://www.wikidata.org/wiki/Q3621696","display_name":"Language model","level":2,"score":0.5239999890327454},{"id":"https://openalex.org/C2776214188","wikidata":"https://www.wikidata.org/wiki/Q408386","display_name":"Inference","level":2,"score":0.47929999232292175},{"id":"https://openalex.org/C11413529","wikidata":"https://www.wikidata.org/wiki/Q8366","display_name":"Algorithm","level":1,"score":0.4659000039100647},{"id":"https://openalex.org/C57273362","wikidata":"https://www.wikidata.org/wiki/Q576722","display_name":"Decoding methods","level":2,"score":0.46140000224113464},{"id":"https://openalex.org/C2779343474","wikidata":"https://www.wikidata.org/wiki/Q3109175","display_name":"Context (archaeology)","level":2,"score":0.4174000024795532},{"id":"https://openalex.org/C117251300","wikidata":"https://www.wikidata.org/wiki/Q1849855","display_name":"Parametric statistics","level":2,"score":0.39640000462532043},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.3702000081539154},{"id":"https://openalex.org/C2780505938","wikidata":"https://www.wikidata.org/wiki/Q17093282","display_name":"Unavailability","level":2,"score":0.3479999899864197},{"id":"https://openalex.org/C155512373","wikidata":"https://www.wikidata.org/wiki/Q287450","display_name":"Residual","level":2,"score":0.3450999855995178},{"id":"https://openalex.org/C63479239","wikidata":"https://www.wikidata.org/wiki/Q7353546","display_name":"Robustness (evolution)","level":3,"score":0.3440000116825104},{"id":"https://openalex.org/C80444323","wikidata":"https://www.wikidata.org/wiki/Q2878974","display_name":"Theoretical computer science","level":1,"score":0.335099995136261},{"id":"https://openalex.org/C28490314","wikidata":"https://www.wikidata.org/wiki/Q189436","display_name":"Speech recognition","level":1,"score":0.3158999979496002},{"id":"https://openalex.org/C176217482","wikidata":"https://www.wikidata.org/wiki/Q860554","display_name":"Metric (unit)","level":2,"score":0.30239999294281006},{"id":"https://openalex.org/C15569618","wikidata":"https://www.wikidata.org/wiki/Q3561421","display_name":"Liveness","level":2,"score":0.2955999970436096},{"id":"https://openalex.org/C2780898871","wikidata":"https://www.wikidata.org/wiki/Q860554","display_name":"Performance metric","level":2,"score":0.2930999994277954},{"id":"https://openalex.org/C99498987","wikidata":"https://www.wikidata.org/wiki/Q2210247","display_name":"Noise (video)","level":3,"score":0.28929999470710754},{"id":"https://openalex.org/C2779843651","wikidata":"https://www.wikidata.org/wiki/Q7390335","display_name":"SIGNAL (programming language)","level":2,"score":0.27950000762939453},{"id":"https://openalex.org/C108154423","wikidata":"https://www.wikidata.org/wiki/Q1469792","display_name":"Salience (neuroscience)","level":2,"score":0.26910001039505005},{"id":"https://openalex.org/C1921717","wikidata":"https://www.wikidata.org/wiki/Q1334846","display_name":"Mahalanobis distance","level":2,"score":0.26330000162124634},{"id":"https://openalex.org/C205606062","wikidata":"https://www.wikidata.org/wiki/Q5249645","display_name":"Decoupling (probability)","level":2,"score":0.2628999948501587},{"id":"https://openalex.org/C179518139","wikidata":"https://www.wikidata.org/wiki/Q5140297","display_name":"Coding (social sciences)","level":2,"score":0.2615000009536743},{"id":"https://openalex.org/C28855332","wikidata":"https://www.wikidata.org/wiki/Q198099","display_name":"Quantization (signal processing)","level":2,"score":0.2590000033378601},{"id":"https://openalex.org/C103088060","wikidata":"https://www.wikidata.org/wiki/Q1062839","display_name":"Error detection and correction","level":2,"score":0.25380000472068787},{"id":"https://openalex.org/C74172769","wikidata":"https://www.wikidata.org/wiki/Q1446839","display_name":"Routing (electronic design automation)","level":2,"score":0.2515999972820282},{"id":"https://openalex.org/C175291020","wikidata":"https://www.wikidata.org/wiki/Q1156822","display_name":"Offset (computer science)","level":2,"score":0.25049999356269836}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.48550/arxiv.2606.01923","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2606.01923","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":null,"license_id":null,"version":null,"is_accepted":false,"is_published":null,"raw_source_name":null,"raw_type":"Preprint"}],"best_oa_location":{"id":"doi:10.48550/arxiv.2606.01923","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2606.01923","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":null,"license_id":null,"version":null,"is_accepted":false,"is_published":false,"raw_source_name":null,"raw_type":"Preprint"},"sustainable_development_goals":[{"score":0.5253720283508301,"display_name":"Quality Education","id":"https://metadata.un.org/sdg/4"}],"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],"frequently":[4],"exhibit":[5],"\"contextual":[6],"disregard\"":[7],"when":[8],"faced":[9],"with":[10,15],"input":[11,159],"evidence":[12,86,160],"that":[13,186,207],"conflicts":[14],"their":[16],"internal":[17],"parametric":[18,202],"memory,":[19],"leading":[20],"to":[21,79,132,141,172],"persistent":[22],"factual":[23,194],"hallucinations.":[24,203],"Existing":[25],"mitigation":[26],"strategies":[27],"primarily":[28],"rely":[29],"on":[30,180],"suppressing":[31,201],"specific":[32],"neuron":[33],"activations":[34],"or":[35,48],"employing":[36],"computationally":[37,212],"expensive":[38],"contrastive":[39],"decoding":[40],"mechanisms,":[41],"which":[42],"often":[43],"result":[44],"in":[45,69,221],"increased":[46],"perplexity":[47],"significantly":[49,188],"elevated":[50],"inference":[51],"latency.":[52],"To":[53],"address":[54],"these":[55],"limitations,":[56],"we":[57,123],"propose":[58],"Resonant":[59],"Context":[60],"Anchoring":[61],"(RCA),":[62],"a":[63,125,209,218],"lightweight":[64],"inference-time":[65],"intervention":[66],"method":[67],"grounded":[68],"the":[70,81,97,107,135,146,154,162,169,173,181,227],"perspective":[71],"of":[72,84,100,120,137,158],"residual":[73,163],"stream":[74,164],"signal":[75,82],"dynamics.":[76],"RCA":[77,187,216],"aims":[78],"resolve":[80],"attenuation":[83],"external":[85],"during":[87,176],"its":[88],"propagation":[89],"through":[90],"deep":[91],"networks.":[92],"The":[93],"core":[94],"mechanism":[95,151],"involves":[96],"orthogonal":[98],"decoupling":[99],"routing":[101],"logic":[102],"and":[103,196,211,223],"information":[104],"magnitude":[105],"within":[106,161],"self-attention":[108],"module.":[109],"By":[110],"utilizing":[111],"raw":[112],"pre-softmax":[113],"attention":[114,147],"scores":[115],"as":[116,208],"an":[117],"instantaneous":[118],"metric":[119],"semantic":[121],"alignment,":[122],"construct":[124],"dynamic":[126],"gain":[127],"field":[128],"via":[129],"non-linear":[130],"rectification":[131],"selectively":[133],"amplify":[134],"norms":[136],"value":[138],"vectors":[139],"corresponding":[140],"context":[142,175],"tokens,":[143],"without":[144],"altering":[145],"probability":[148],"distribution.":[149],"This":[150],"effectively":[152,200],"elevates":[153],"signal-to-noise":[155],"ratio":[156],"(SNR)":[157],"mixture,":[165],"thereby":[166],"robustly":[167],"anchoring":[168],"generation":[170],"trajectory":[171],"truthful":[174],"inference.":[177],"Extensive":[178],"experiments":[179],"Llama-3":[182],"model":[183],"series":[184],"demonstrate":[185],"improves":[189],"contextual":[190],"faithfulness":[191,222],"across":[192],"multiple":[193],"consistency":[195],"strong":[197],"knowledge-conflict":[198],"tasks,":[199],"Furthermore,":[204],"results":[205],"confirm":[206],"training-free":[210],"negligible":[213],"plug-and-play":[214],"module,":[215],"achieves":[217],"Pareto":[219],"improvement":[220],"fluency":[224],"while":[225],"maintaining":[226],"model's":[228],"general":[229],"language":[230],"understanding":[231],"capabilities.":[232]},"counts_by_year":[],"updated_date":"2026-07-01T06:00:48.157686","created_date":"2026-06-03T00:00:00"}