{"id":"W4392484188","doi":"10.1145/3636555.3636910","title":"Prompt-based and Fine-tuned GPT Models for Context-Dependent and -Independent Deductive Coding in Social Annotation","year":2024,"lang":"en","type":"article","venue":"","topic":"Innovative Teaching and Learning Methods","field":"Psychology","cited_by":31,"is_retracted":false,"has_abstract":true,"ca_institutions":"McGill University","funders":"","keywords":"Computer science; Coding (social sciences); Annotation; Natural language processing; Artificial intelligence; Machine learning; Mathematics","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":[0.0142433,0.001393858,0.001053127,0.001666772,0.001011834,0.002788853,0.002894523,0.002008867,0.004437682],"category_scores_gemma":[0.07917274,0.0008688734,0.001685564,0.001225246,0.001769142,0.004788205,0.004128018,0.004537861,0.002006078],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.003523172,"about_ca_system_score_gemma":0.003771425,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.008762726,"about_ca_topic_score_gemma":0.01188977,"domain_scores_codex":[0.9921677,0.004222633,0.0005016175,0.001899548,0.0009332616,0.0002751716],"domain_scores_gemma":[0.9485837,0.03834266,0.001806387,0.004658428,0.005810559,0.0007982681],"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":[0.0010535,0.0005957379,0.01606068,0.0008064493,0.000171406,0.0005683189,0.007411801,0.4477642,0.01126051,0.04200066,0.007740445,0.4645663],"study_design_scores_gemma":[0.00003950793,0.00007958234,0.0006172099,0.00004626926,0.00002414986,0.00005412731,0.00026892,0.9617617,0.002393869,0.03231728,0.002364,0.00003332344],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.02296372,0.0001010153,0.9714032,0.0003864289,0.00007887485,0.0004824612,0.0004361328,0.003007658,0.001140522],"genre_scores_gemma":[0.3331266,0.0001033078,0.6614728,0.0003069422,0.00004280162,0.001457365,0.001413965,0.0003337152,0.001742476],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.0142433,"threshold_uncertainty_score":0.07532668,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.08746716021728913,"score_gpt":0.4109897229639487,"score_spread":0.3235225627466596,"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."}}