{"id":"W4386763311","doi":"10.36227/techrxiv.24143706.v1","title":"GPT-4 as a Twitter Data Annotator: Unraveling Its Performance on a Stance Classification Task","year":2023,"lang":"en","type":"preprint","venue":"","topic":"Topic Modeling","field":"Computer Science","cited_by":6,"is_retracted":false,"has_abstract":true,"ca_institutions":"Lakehead University; York University","funders":"Social Sciences and Humanities Research Council of Canada; Natural Sciences and Engineering Research Council of Canada","keywords":"Task (project management); Computer science; Natural language processing; Artificial intelligence; Engineering","routes":{"ca_aff":true,"ca_fund":true,"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","insufficient_payload"],"consensus_categories":[],"category_scores_codex":[0.0008579449,0.000364293,0.0003213838,0.0002419386,0.0001460956,0.0004811377,0.005055072,0.0002600496,0.00002756891],"category_scores_gemma":[0.0001309454,0.0003453752,0.00006163886,0.000319843,0.00002483391,0.0005483163,0.005004492,0.0008666457,0.001536715],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0001598398,"about_ca_system_score_gemma":0.0003922994,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.00006014049,"about_ca_topic_score_gemma":0.0000118532,"domain_scores_codex":[0.9961196,0.00007962451,0.0005706012,0.001985923,0.0007779632,0.0004662596],"domain_scores_gemma":[0.9941691,0.0001145614,0.0002930606,0.005142545,0.0001780298,0.0001027557],"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":[0.000169951,0.0006738353,0.004704811,0.003989784,0.0006423069,0.0002713754,0.01811796,0.1981357,0.005195102,0.2521699,0.1163237,0.3996055],"study_design_scores_gemma":[0.0001369561,0.00002830537,0.001095306,0.000357555,0.000008622481,0.000004316443,0.0000424275,0.9886779,0.0003326701,0.002194895,0.006711042,0.0004099813],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.05224184,0.0001340348,0.9312309,0.004926294,0.002611597,0.0005877809,0.00003509379,0.001074952,0.0071575],"genre_scores_gemma":[0.9269764,0.0003177782,0.06402609,0.001830784,0.0004743506,0.00009407127,0.0001681549,0.00005332024,0.006059012],"genre_candidate":"methods","genre_consensus":null,"teacher_disagreement_score":0.8747346,"threshold_uncertainty_score":0.9998998,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.2535944618526218,"score_gpt":0.3486720213166273,"score_spread":0.09507755946400553,"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."}}