{"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":"metacan-v3-hybrid-931329e0061c","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.001055883,0.001699217,0.0008816162,0.0006269484,0.0005508639,0.001298452,0.002852804,0.0009597784,0.007159622],"category_scores_gemma":[0.004711397,0.0007181536,0.001309929,0.0005237352,0.0007395049,0.002372772,0.002354034,0.001769489,0.002113155],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.000892659,"about_ca_system_score_gemma":0.00226385,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.007613067,"about_ca_topic_score_gemma":0.02065659,"domain_scores_codex":[0.9992228,0.0002131071,0.00004608298,0.0001518098,0.0002252305,0.0001409471],"domain_scores_gemma":[0.9989477,0.0003356451,0.0001018131,0.0003716793,0.0001612119,0.00008193485],"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":[0.0008088215,0.0004859542,0.01123894,0.0006495914,0.0003402605,0.0006552322,0.0004549222,0.6210372,0.02887574,0.02572514,0.04529672,0.2644314],"study_design_scores_gemma":[0.00003779463,0.00004175958,0.000216715,0.0000168376,0.00001765334,0.0000461342,0.00005092521,0.9862052,0.003429312,0.005862535,0.004061235,0.00001396539],"study_design_candidate":"simulation_or_modeling","study_design_consensus":"simulation_or_modeling","genre_codex":"methods","genre_gemma":"methods","genre_scores_codex":[0.05581008,0.001156796,0.8662578,0.0005665342,0.000261871,0.0001255455,0.001055329,0.06369869,0.01106729],"genre_scores_gemma":[0.5290786,0.0003011186,0.4548153,0.0005802824,0.00006252675,0.0002164549,0.003652996,0.005678128,0.005614633],"genre_candidate":"methods","genre_consensus":"methods","teacher_disagreement_score":0.007613067,"threshold_uncertainty_score":0.02395129,"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."}}