{"id":"W4396650996","doi":"10.1145/3640457.3688142","title":"Bayesian Optimization with LLM-Based Acquisition Functions for Natural Language Preference Elicitation","year":2024,"lang":"en","type":"preprint","venue":"","topic":"Speech and dialogue systems","field":"Computer Science","cited_by":5,"is_retracted":false,"has_abstract":true,"ca_institutions":"University of Toronto; University of Waterloo","funders":"","keywords":"Preference elicitation; Preference; Computer science; Bayesian optimization; Natural language; Natural (archaeology); Artificial intelligence; Natural language processing; Mathematics; Statistics; Biology","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":[],"consensus_categories":[],"category_scores_codex":[0.0001873935,0.000237704,0.0001965439,0.0002682946,0.00011178,0.0005446535,0.0004008396,0.0001824801,0.00002340664],"category_scores_gemma":[0.00003132335,0.0001864336,0.0001032448,0.0003092237,0.00002089879,0.0001949958,0.0001856596,0.0002469966,0.00003868273],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0001186041,"about_ca_system_score_gemma":0.0002827279,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.0001593386,"about_ca_topic_score_gemma":0.0001220209,"domain_scores_codex":[0.9985248,0.00005610222,0.0002474792,0.0006628975,0.0002843564,0.0002243883],"domain_scores_gemma":[0.9988644,0.0001061592,0.0001351381,0.0005658229,0.0002545828,0.00007389563],"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":[0.0002300161,0.0002143133,0.0000570351,0.00159785,0.0001849745,0.0000193774,0.002022965,0.931696,0.0007070214,0.02197821,0.0112204,0.03007187],"study_design_scores_gemma":[0.0003458927,0.0001591165,0.00007709354,0.0002265363,0.00003623071,0.000004329685,0.00009487633,0.9959587,0.0008531497,0.001902546,0.00006912363,0.0002724054],"study_design_candidate":"simulation_or_modeling","study_design_consensus":"simulation_or_modeling","genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.0007097144,0.0003786049,0.9917347,0.0007105924,0.001902987,0.001067976,0.00007316371,0.0007072936,0.002714987],"genre_scores_gemma":[0.7697031,0.000001545354,0.2274791,0.0002052101,0.0002496556,0.0004027064,0.001093495,0.00002051617,0.0008447225],"genre_candidate":"methods","genre_consensus":null,"teacher_disagreement_score":0.7689934,"threshold_uncertainty_score":0.760254,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.01651333121361138,"score_gpt":0.2480529212065353,"score_spread":0.231539589992924,"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."}}