{"id":"W4409537504","doi":"10.1145/3727200.3727220","title":"Towards Sustainable Large Language Model Serving","year":2024,"lang":"en","type":"article","venue":"ACM SIGEnergy Energy Informatics Review","topic":"Natural Language Processing Techniques","field":"Computer Science","cited_by":13,"is_retracted":false,"has_abstract":true,"ca_institutions":"University of Waterloo","funders":"","keywords":"Computer science; Business","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":[0.003122073,0.0006782564,0.0008371837,0.0007720403,0.0006849624,0.003073263,0.002627941,0.001627909,0.00503302],"category_scores_gemma":[0.009593454,0.000528749,0.001497653,0.001212853,0.001530858,0.00687746,0.00277729,0.003039563,0.002125426],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.002570883,"about_ca_system_score_gemma":0.002589793,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.007562078,"about_ca_topic_score_gemma":0.01190204,"domain_scores_codex":[0.997968,0.0009115716,0.00009557649,0.0002455561,0.0006411016,0.0001382031],"domain_scores_gemma":[0.9949204,0.003203681,0.0001894477,0.0009138547,0.0006613281,0.000111299],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"theoretical_or_conceptual","study_design_gemma":"not_applicable","study_design_scores_codex":[0.00008955666,0.0001379869,0.0009420353,0.0004982433,0.0001008439,0.0003111027,0.0007370001,0.3598594,0.005359955,0.490353,0.03119518,0.1104156],"study_design_scores_gemma":[0.00001005676,0.00001290027,0.00006206689,0.00002274067,0.000009710683,0.00005676198,0.0001313094,0.7871527,0.001513606,0.1901562,0.02085954,0.00001237417],"study_design_candidate":"not_applicable","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.01210559,0.001149084,0.9678994,0.004794863,0.0001324279,0.00008785014,0.0007323942,0.004277939,0.008820311],"genre_scores_gemma":[0.2620654,0.00190799,0.7141131,0.002040323,0.0002844947,0.0003846402,0.003192748,0.002608309,0.01340298],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.007562078,"threshold_uncertainty_score":0.01865315,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.01200015808108999,"score_gpt":0.290577730194465,"score_spread":0.278577572113375,"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."}}