{"id":"W4416541694","doi":"10.48550/arxiv.2504.11814","title":"ARWI: Arabic Write and Improve","year":2025,"lang":"en","type":"preprint","venue":"ArXiv.org","topic":"Text Readability and Simplification","field":"Computer Science","cited_by":0,"is_retracted":false,"has_abstract":true,"ca_institutions":"","funders":"York University; New York University Abu Dhabi","keywords":"Arabic; Grammar; Modern Standard Arabic; Profiling (computer programming); Arabic languages; Error analysis","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":"codex-gemma-dda1882f352a","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.0003217547,0.0002250514,0.0002698873,0.0001071458,0.0001113653,0.0001823283,0.0009770797,0.0002666334,0.000007560207],"category_scores_gemma":[0.0001007218,0.0002175608,0.00009229827,0.0001696844,0.00008197187,0.0002156178,0.001563867,0.0005283536,0.00006100983],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.00006310624,"about_ca_system_score_gemma":0.0001820088,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.0001361291,"about_ca_topic_score_gemma":0.00002643761,"domain_scores_codex":[0.9982995,0.00007693615,0.0003112569,0.0009092907,0.0001624358,0.0002405858],"domain_scores_gemma":[0.9980381,0.0001287462,0.0001409152,0.001495039,0.0001061859,0.00009105507],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"observational","study_design_gemma":"observational","study_design_scores_codex":[0.00002293786,0.0003282288,0.534847,0.001217771,0.0001602351,0.00002049775,0.003329032,0.0001557204,0.004884728,0.06445312,0.001453538,0.3891273],"study_design_scores_gemma":[0.0002725497,0.00005597845,0.9398832,0.0001665443,0.00003641577,0.000006106827,0.00003246468,0.01106532,0.003381394,0.03089753,0.01361762,0.0005848745],"study_design_candidate":"observational","study_design_consensus":"observational","genre_codex":"empirical","genre_gemma":"empirical","genre_scores_codex":[0.8774134,0.0006955949,0.1132461,0.003848525,0.001218376,0.0004694225,0.00002513375,0.0003166888,0.002766822],"genre_scores_gemma":[0.9928493,0.0001374995,0.004428046,0.0004825626,0.0001144497,0.0000636823,0.0000164996,0.000006561044,0.001901383],"genre_candidate":"empirical","genre_consensus":"empirical","teacher_disagreement_score":0.4050362,"threshold_uncertainty_score":0.887187,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.03535004254881862,"score_gpt":0.2728030644487988,"score_spread":0.2374530218999802,"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."}}