{"id":"W4415775588","doi":"10.1101/2025.07.18.664723","title":"Identifying intervention strategies from machine learning models with COALA: a counterfactual optimization framework","year":2025,"lang":"en","type":"preprint","venue":"bioRxiv (Cold Spring Harbor Laboratory)","topic":"Explainable Artificial Intelligence (XAI)","field":"Computer Science","cited_by":0,"is_retracted":false,"has_abstract":true,"ca_institutions":"Queen's University","funders":"","keywords":"Interpretability; Counterfactual thinking; Counterfactual conditional; Feature (linguistics); Psychological intervention; Replicate; Focus (optics); Biomedicine","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"],"consensus_categories":[],"category_scores_codex":[0.000702166,0.000674938,0.0006282978,0.0004651899,0.0004011149,0.002550876,0.001846431,0.0005854256,0.00005268094],"category_scores_gemma":[0.0001977169,0.0007155064,0.0001717927,0.0009243264,0.0001222086,0.002018949,0.001428393,0.001635179,0.00003664607],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0004135062,"about_ca_system_score_gemma":0.0006842287,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.00112996,"about_ca_topic_score_gemma":0.00004805015,"domain_scores_codex":[0.9959997,0.0003327898,0.0007909343,0.001561829,0.0006790862,0.0006356693],"domain_scores_gemma":[0.9964907,0.0002384447,0.0007124993,0.001567314,0.0008099118,0.0001811478],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"simulation_or_modeling","study_design_gemma":"simulation_or_modeling","study_design_scores_codex":[0.00006007984,0.0001808986,0.0007470313,0.0003345234,0.0002772025,0.00007259053,0.0003410889,0.9469586,0.005318799,0.04566899,0.0000229726,0.00001724812],"study_design_scores_gemma":[0.000223626,0.00009945729,0.0003558257,0.002588895,0.0000984382,2.000173e-8,0.00008818346,0.9470698,0.04767761,0.0008124783,0.0001137521,0.0008719497],"study_design_candidate":"simulation_or_modeling","study_design_consensus":"simulation_or_modeling","genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.04952882,0.001219036,0.9463037,0.0001324488,0.001143538,0.0005758218,0.00009034274,0.0009766691,0.00002961706],"genre_scores_gemma":[0.7779979,0.0002165525,0.2213632,0.00007937488,0.0001415566,0.0001327243,0.000001714624,0.00005626826,0.00001067795],"genre_candidate":"methods","genre_consensus":null,"teacher_disagreement_score":0.7284691,"threshold_uncertainty_score":0.9995296,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.03052836204450768,"score_gpt":0.2541855322145841,"score_spread":0.2236571701700764,"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."}}