{"id":"W4362577036","doi":"10.22215/etd/2023-15442","title":"MAWT: Multi-Attention-Weight Transformers","year":2023,"lang":"en","type":"dissertation","venue":"","topic":"Natural Language Processing Techniques","field":"Computer Science","cited_by":0,"is_retracted":false,"has_abstract":true,"ca_institutions":"Carleton University","funders":"Ministère de la Défense Nationale","keywords":"Transformer; Computer science; Machine translation; Artificial intelligence; Parsing; Test set; Natural language processing; Artificial neural network; Machine learning; Engineering; Voltage","routes":{"ca_aff":true,"ca_fund":true,"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":[0.001115735,0.001274371,0.0006164443,0.0008866299,0.0004611436,0.001421421,0.002979893,0.001127735,0.01191529],"category_scores_gemma":[0.005034704,0.000584765,0.001204326,0.0007823323,0.000701979,0.004727846,0.001726785,0.002071905,0.005714698],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.000834378,"about_ca_system_score_gemma":0.001139737,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.003012604,"about_ca_topic_score_gemma":0.005051649,"domain_scores_codex":[0.9993332,0.0001184162,0.00007073706,0.0002142343,0.0001828861,0.00008053549],"domain_scores_gemma":[0.9988278,0.0004483846,0.00009256235,0.0003354508,0.0002379256,0.00005804115],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"design_other","study_design_gemma":"bench_or_experimental","study_design_scores_codex":[0.0006932221,0.000260592,0.003900195,0.0003688584,0.0001871488,0.0005083536,0.0003144548,0.08145485,0.03305625,0.0495805,0.03695402,0.7927216],"study_design_scores_gemma":[0.0000716227,0.000235958,0.0005776642,0.00004258934,0.0001142632,0.0003710211,0.0000528669,0.8695788,0.03582012,0.0731248,0.0199632,0.00004706305],"study_design_candidate":"bench_or_experimental","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.01983388,0.0004469286,0.943117,0.0003277636,0.0002740145,0.0002143855,0.0008312367,0.02897444,0.00598046],"genre_scores_gemma":[0.528033,0.0005730269,0.4385659,0.0008491143,0.0001733274,0.0005392431,0.00352085,0.002701314,0.02504427],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.01191529,"threshold_uncertainty_score":0.03986067,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.01419145913062106,"score_gpt":0.2986226598812279,"score_spread":0.2844312007506068,"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."}}