{"id":"W4399447660","doi":"10.48550/arxiv.2406.02969","title":"Filtered not Mixed: Stochastic Filtering-Based Online Gating for Mixture of Large Language Models","year":2024,"lang":"en","type":"preprint","venue":"arXiv (Cornell University)","topic":"Topic Modeling","field":"Computer Science","cited_by":0,"is_retracted":false,"has_abstract":true,"ca_institutions":"","funders":"Vector Institute; McMaster University; Natural Sciences and Engineering Research Council of Canada; Government of Canada; Canadian Institute for Advanced Research","keywords":"Computer science; Gating; Language model; Artificial intelligence; Psychology; Neuroscience","routes":{"ca_aff":false,"ca_fund":true,"ca_venue":false,"about_ca":false,"invisible_to_affiliation_only":true},"retraction":null,"screen":null,"direct_labels":[],"prediction":{"model_version":"metacan-v3-hybrid-931329e0061c","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.003643121,0.001430234,0.001698498,0.000957866,0.0007449971,0.001649964,0.003013783,0.001818622,0.00326939],"category_scores_gemma":[0.009344738,0.0009782849,0.001248407,0.001049096,0.001017159,0.003563224,0.002860575,0.002565559,0.0009996878],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.001197782,"about_ca_system_score_gemma":0.002073117,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.007052794,"about_ca_topic_score_gemma":0.01008449,"domain_scores_codex":[0.9984927,0.0004934092,0.00007464682,0.0003388911,0.0003767454,0.0002236057],"domain_scores_gemma":[0.9969919,0.002029748,0.0001863782,0.0003887932,0.0002520666,0.0001511064],"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":[0.0003805629,0.0002534869,0.002460778,0.0001037558,0.0001719998,0.0001924111,0.0002485465,0.5889332,0.007694002,0.0631213,0.004391225,0.3320488],"study_design_scores_gemma":[0.000009463572,0.00002256775,0.00008508067,0.000005589997,0.000009355853,0.00001689663,0.000005435581,0.9868327,0.0009732204,0.0113195,0.0007101929,0.00001009905],"study_design_candidate":"simulation_or_modeling","study_design_consensus":"simulation_or_modeling","genre_codex":"methods","genre_gemma":"methods","genre_scores_codex":[0.004572059,0.0001250988,0.9937412,0.0001114131,0.00002450604,0.00002736379,0.00004779794,0.0008182512,0.0005323322],"genre_scores_gemma":[0.3972564,0.0003498546,0.5965911,0.0006388196,0.0001789846,0.0002863235,0.000598667,0.0004415574,0.003658184],"genre_candidate":"methods","genre_consensus":"methods","teacher_disagreement_score":0.007052794,"threshold_uncertainty_score":0.0192669,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.08123969319778795,"score_gpt":0.2250849381751377,"score_spread":0.1438452449773497,"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."}}