{"id":"W4407677752","doi":"10.2196/70014","title":"Evaluating Generative AI in Mental Health: Systematic Review of Capabilities and Limitations","year":2025,"lang":"en","type":"review","venue":"JMIR Mental Health","topic":"Digital Mental Health Interventions","field":"Psychology","cited_by":41,"is_retracted":false,"has_abstract":true,"ca_institutions":"","funders":"","keywords":"Preprint; Mental health; Psychology; Artificial intelligence; Computer science; Psychiatry; World Wide Web","routes":{"ca_aff":false,"ca_fund":false,"ca_venue":true,"about_ca":false,"invisible_to_affiliation_only":true},"retraction":null,"screen":null,"direct_labels":[],"prediction":{"model_version":"metacan-v3-hybrid-931329e0061c","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.04072275,0.001698597,0.007499964,0.01199766,0.001000297,0.005232313,0.003372135,0.002211161,0.005433716],"category_scores_gemma":[0.1966377,0.001661607,0.01032742,0.01111411,0.002396464,0.004602078,0.003912888,0.001992735,0.0004232526],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.006074021,"about_ca_system_score_gemma":0.02411959,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.01285429,"about_ca_topic_score_gemma":0.03830178,"domain_scores_codex":[0.9683332,0.01417415,0.01172449,0.001318118,0.004032805,0.0004172702],"domain_scores_gemma":[0.7858461,0.185399,0.0153287,0.003185713,0.00932458,0.0009160439],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"systematic_review","study_design_gemma":"systematic_review","study_design_scores_codex":[0.0001588047,0.00002000524,0.0005639099,0.9242423,0.005855504,0.00005267308,0.000500304,0.0001632123,0.00006291206,0.0005695261,0.0009178173,0.06689312],"study_design_scores_gemma":[0.0001394978,0.0001529951,0.001275971,0.9599534,0.02414435,0.0001219602,0.0004206547,0.00008687379,0.0001091515,0.0006470623,0.01291543,0.00003258351],"study_design_candidate":"systematic_review","study_design_consensus":"systematic_review","genre_codex":"review","genre_gemma":"review","genre_scores_codex":[0.001115557,0.9959382,0.0005595365,0.000499117,0.0001237458,0.0007257644,0.0004334111,0.00001776912,0.0005868477],"genre_scores_gemma":[0.02355072,0.969961,0.003029063,0.000865325,0.00007495833,0.001990169,0.000407584,0.00001827822,0.0001027801],"genre_candidate":"review","genre_consensus":"review","teacher_disagreement_score":0.04072275,"threshold_uncertainty_score":0.2153651,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.2092849492106648,"score_gpt":0.5638544977107439,"score_spread":0.3545695485000791,"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."}}