{"id":"W3135618437","doi":"10.21203/rs.3.rs-246079/v1","title":"Towards AI-powered Language Assessment Tools","year":2021,"lang":"en","type":"preprint","venue":"Research Square","topic":"Natural Language Processing Techniques","field":"Computer Science","cited_by":5,"is_retracted":false,"has_abstract":false,"ca_institutions":"University of Toronto","funders":"","keywords":"Computer science; Natural language processing","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":[0.01064141,0.001190003,0.0008993461,0.004460678,0.0008755908,0.009379179,0.002937417,0.002263386,0.01187469],"category_scores_gemma":[0.04548597,0.0007734874,0.0009187392,0.002717211,0.002561008,0.01239975,0.005508796,0.004011787,0.006670319],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.001164493,"about_ca_system_score_gemma":0.002657415,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.001146908,"about_ca_topic_score_gemma":0.001326732,"domain_scores_codex":[0.989543,0.005338091,0.000628797,0.000866592,0.003336461,0.0002871004],"domain_scores_gemma":[0.9701081,0.01717896,0.001005293,0.004690252,0.00617044,0.0008470487],"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":[0.0001570859,0.0003704228,0.001943472,0.0006322681,0.00008657163,0.0002520256,0.001399608,0.009455042,0.01428472,0.428256,0.01775907,0.5254037],"study_design_scores_gemma":[0.00005217043,0.00008211993,0.0006135364,0.0003525892,0.00005005358,0.0002293182,0.000692625,0.1707741,0.01795915,0.7333857,0.07575586,0.00005279824],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"methods","genre_scores_codex":[0.003851741,0.0003045316,0.9775474,0.001108933,0.00008955861,0.0001543237,0.0001567994,0.004835093,0.01195152],"genre_scores_gemma":[0.09828513,0.0004986987,0.8884151,0.0005260758,0.0001715966,0.0004274938,0.0007882447,0.0009883041,0.009899395],"genre_candidate":"methods","genre_consensus":"methods","teacher_disagreement_score":0.01187469,"threshold_uncertainty_score":0.05627781,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.0667866618727071,"score_gpt":0.4603486543558466,"score_spread":0.3935619924831395,"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."}}