{"id":"W4404752327","doi":"10.18653/v1/2022.aacl-main.39","title":"Director: Generator-Classifiers For Supervised Language Modeling","year":2022,"lang":"en","type":"article","venue":"","topic":"Natural Language Processing Techniques","field":"Computer Science","cited_by":11,"is_retracted":false,"has_abstract":true,"ca_institutions":"McGill University","funders":"","keywords":"Computer science; Generator (circuit theory); Artificial intelligence; Language model; Natural language processing; Physics; Power (physics)","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":[0.003512885,0.001844139,0.001520689,0.002295235,0.001114339,0.002237651,0.003662699,0.002458113,0.02070516],"category_scores_gemma":[0.01162427,0.001588493,0.001627111,0.00142455,0.0006892508,0.004883482,0.00292248,0.003908469,0.02650728],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0008035719,"about_ca_system_score_gemma":0.001725674,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.00372197,"about_ca_topic_score_gemma":0.008749693,"domain_scores_codex":[0.9976572,0.0009222451,0.0001309526,0.0006249695,0.0005229895,0.0001417308],"domain_scores_gemma":[0.9962675,0.001885533,0.0001103939,0.0009859388,0.0005941542,0.0001564187],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"design_other","study_design_gemma":"simulation_or_modeling","study_design_scores_codex":[0.0007080663,0.0002250605,0.001659037,0.0003171913,0.0002389144,0.0001683859,0.0001328901,0.03091448,0.004207788,0.02011867,0.3771462,0.5641633],"study_design_scores_gemma":[0.000238496,0.00009006234,0.0004322552,0.00005685028,0.00006368247,0.0001476139,0.0000488392,0.8666865,0.01119848,0.0595682,0.06142498,0.00004395454],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"methods","genre_scores_codex":[0.002082816,0.0007647017,0.8979804,0.0005343638,0.0004278253,0.000161012,0.004911656,0.09025554,0.002881624],"genre_scores_gemma":[0.08522369,0.0008581823,0.8390605,0.0009311634,0.0006496214,0.001169132,0.03733953,0.009749952,0.02501828],"genre_candidate":"methods","genre_consensus":"methods","teacher_disagreement_score":0.02070516,"threshold_uncertainty_score":0.06926566,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.02489321284759761,"score_gpt":0.2742018675418575,"score_spread":0.2493086546942599,"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."}}