{"id":"W7123868952","doi":"10.1145/3772052.3772215","title":"Oneiros: KV Cache Optimization through Parameter Remapping for Multi-tenant LLM Serving","year":2025,"lang":"","type":"article","venue":"","topic":"Parallel Computing and Optimization Techniques","field":"Computer Science","cited_by":1,"is_retracted":false,"has_abstract":true,"ca_institutions":"Becton Dickinson (Canada)","funders":"National Science Foundation","keywords":"Cache; CPU cache; Key (lock); Throughput; Cache coloring; Central processing unit; Cache pollution; Memory bandwidth","routes":{"ca_aff":true,"ca_fund":false,"ca_venue":false,"about_ca":false,"invisible_to_affiliation_only":false},"retraction":null,"screen":null,"direct_labels":[],"prediction":{"model_version":"codex-gemma-dda1882f352a","candidate_categories":["metaepi_narrow","scholarly_communication"],"consensus_categories":[],"category_scores_codex":[0.00115929,0.0006017276,0.0007172781,0.0003548921,0.001040161,0.00108084,0.001682319,0.0004368853,0.00004173558],"category_scores_gemma":[0.0006668573,0.0005044083,0.000368744,0.00158515,0.0001258467,0.001166954,0.001033591,0.0003559264,0.00001301277],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0002393607,"about_ca_system_score_gemma":0.0003857591,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.0001524155,"about_ca_topic_score_gemma":0.00001228228,"domain_scores_codex":[0.9956272,0.0003012976,0.001255051,0.0014891,0.0003486541,0.0009787278],"domain_scores_gemma":[0.9966716,0.0005802581,0.0004803315,0.001342908,0.0007941173,0.0001308388],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"simulation_or_modeling","study_design_gemma":"simulation_or_modeling","study_design_scores_codex":[0.00005129404,0.0005902224,0.0002191809,0.000427292,0.000192813,0.000003260225,0.001860392,0.9274341,0.0001325705,0.02635997,0.00814966,0.0345792],"study_design_scores_gemma":[0.0009479878,0.000212613,0.00005461865,0.0007551454,0.00006980586,0.000002783083,0.0001074623,0.9846165,0.004850603,0.001947346,0.005772884,0.0006623054],"study_design_candidate":"simulation_or_modeling","study_design_consensus":"simulation_or_modeling","genre_codex":"methods","genre_gemma":"methods","genre_scores_codex":[0.00006827724,0.0009337884,0.9881477,0.003785674,0.001654847,0.001714017,0.000008727918,0.001199192,0.002487792],"genre_scores_gemma":[0.01495467,0.0006382558,0.9698811,0.002900965,0.0001192262,0.0001520796,0.00002777032,0.00004125337,0.01128464],"genre_candidate":"methods","genre_consensus":"methods","teacher_disagreement_score":0.0571823,"threshold_uncertainty_score":0.9999561,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.06892126961872165,"score_gpt":0.3331146934264618,"score_spread":0.2641934238077401,"validation_status":"score_only:v0-immature-baseline","note":"Baseline scores from an immature model (maturity gate not passed). Scores rank; they never assert a category."}}