{"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":"codex-gemma-dda1882f352a","candidate_categories":["metaepi_narrow","scholarly_communication","open_science","research_integrity"],"consensus_categories":[],"category_scores_codex":[0.002876069,0.0003979479,0.0005417378,0.0006172609,0.0002716767,0.004154816,0.004441523,0.0005880302,0.0002086385],"category_scores_gemma":[0.0009946101,0.0003601174,0.0002728072,0.001131255,0.0001358628,0.0007856617,0.01179268,0.004442766,0.00003764706],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0007668764,"about_ca_system_score_gemma":0.002791805,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.000554411,"about_ca_topic_score_gemma":0.00005227071,"domain_scores_codex":[0.9934183,0.0009826433,0.0004319177,0.001445112,0.002696079,0.001025909],"domain_scores_gemma":[0.9949229,0.0002997807,0.0001402878,0.002853645,0.001497488,0.0002859124],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"design_other","study_design_gemma":"theoretical_or_conceptual","study_design_scores_codex":[0.00002401455,0.000664279,0.0004058778,0.004468989,0.0002204229,0.005354792,0.0091577,0.00003227656,0.01386031,0.1299479,0.01706152,0.8188019],"study_design_scores_gemma":[0.002503946,0.00175537,0.01270835,0.02148341,0.00009061287,0.000388288,0.006755465,0.06169046,0.407704,0.4529232,0.02432121,0.007675627],"study_design_candidate":"design_other","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.0068562,0.03317467,0.9277973,0.01139224,0.001048931,0.002059459,0.0001096793,0.002793574,0.01476799],"genre_scores_gemma":[0.5323674,0.0002518204,0.4657478,0.0002731592,0.000261625,0.0003153526,0.0001896275,0.00004501936,0.0005481321],"genre_candidate":"methods","genre_consensus":null,"teacher_disagreement_score":0.8111263,"threshold_uncertainty_score":0.9998851,"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."}}