{"id":"W4389519117","doi":"10.18653/v1/2023.emnlp-main.491","title":"Don’t Trust ChatGPT when your Question is not in English: A Study of Multilingual Abilities and Types of LLMs","year":2023,"lang":"en","type":"article","venue":"","topic":"Topic Modeling","field":"Computer Science","cited_by":62,"is_retracted":false,"has_abstract":true,"ca_institutions":"University of Alberta","funders":"Natural Sciences and Engineering Research Council of Canada; Alberta Machine Intelligence Institute","keywords":"Variety (cybernetics); Computer science; Natural language processing; Artificial intelligence; Linguistics","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":[],"consensus_categories":[],"category_scores_codex":[0.0004035572,0.00006512273,0.0001536223,0.0001413431,0.00001513854,0.00001831341,0.0002114174,0.00003374254,0.00001530722],"category_scores_gemma":[0.0001377599,0.0000570971,0.00001543471,0.0001645236,0.00002355837,0.0001701756,0.0001828963,0.00005489808,0.000001519024],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.00001142659,"about_ca_system_score_gemma":0.00002450395,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.002088544,"about_ca_topic_score_gemma":0.0005067937,"domain_scores_codex":[0.9991926,0.00004751576,0.0002489714,0.0002256351,0.0001761002,0.0001091584],"domain_scores_gemma":[0.9994826,0.00008744992,0.00004932581,0.0002742207,0.00008479862,0.00002154486],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"qualitative","study_design_gemma":"simulation_or_modeling","study_design_scores_codex":[0.00003659783,0.0004643754,0.1708974,0.0001906298,0.00003336845,0.0000084669,0.7517956,0.002519826,0.00115948,0.009751223,0.00007759567,0.06306534],"study_design_scores_gemma":[0.00195031,0.000591996,0.1103124,0.00009749218,0.00001183684,0.000001349766,0.04237564,0.828115,0.01221076,0.003956518,0.0000992659,0.0002774558],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"empirical","genre_gemma":"empirical","genre_scores_codex":[0.9966191,0.00002620022,0.002569542,0.0001079792,0.0001320591,0.0001643495,0.000001339761,0.00007966116,0.0002997686],"genre_scores_gemma":[0.9949059,0.000006206526,0.004551709,0.00002034605,0.0000238607,0.000006303158,2.675278e-7,0.000003438881,0.000481928],"genre_candidate":"empirical","genre_consensus":"empirical","teacher_disagreement_score":0.8255952,"threshold_uncertainty_score":0.3157268,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.04011867324513748,"score_gpt":0.2954704506340845,"score_spread":0.255351777388947,"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."}}