{"id":"W6966502134","doi":"10.48448/azcp-f870","title":"Scope Ambiguities in Large Language Models","year":2024,"lang":"en","type":"other","venue":"Open MIND","topic":"","field":"","cited_by":0,"is_retracted":false,"has_abstract":true,"ca_institutions":"McGill University; National Research Council Canada; Mila - Quebec Artificial Intelligence Institute","funders":"","keywords":"Scope (computer science); Ambiguity; Meaning (existential); Language model; Language understanding; Semantics (computer science)","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":[0.006754641,0.001181019,0.0009591395,0.001697621,0.0008147341,0.002972561,0.001200382,0.001101656,0.002329544],"category_scores_gemma":[0.03323497,0.0007507936,0.001418656,0.001514348,0.001575183,0.004942512,0.001871486,0.002504685,0.000736882],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.001400434,"about_ca_system_score_gemma":0.001128423,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.009353,"about_ca_topic_score_gemma":0.01402012,"domain_scores_codex":[0.9950708,0.003029261,0.0002673459,0.0008868722,0.0006258335,0.0001198316],"domain_scores_gemma":[0.9687858,0.02729976,0.001437405,0.001568317,0.000655536,0.0002532576],"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":[0.000408244,0.0001456605,0.013377,0.0004934571,0.0004157634,0.0007970746,0.002452141,0.649007,0.002844374,0.1809801,0.0116407,0.1374385],"study_design_scores_gemma":[0.00002087925,0.00001785818,0.0007508018,0.00003250873,0.00002201168,0.00008677861,0.0001481899,0.8374375,0.0004913263,0.1586709,0.002296185,0.00002517495],"study_design_candidate":"simulation_or_modeling","study_design_consensus":"simulation_or_modeling","genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.09614904,0.001884668,0.8889371,0.001960975,0.00009363989,0.0001115132,0.001764005,0.003090521,0.006008473],"genre_scores_gemma":[0.8309171,0.0008031575,0.1614963,0.0005080473,0.0001552427,0.0002006722,0.00317418,0.0005619094,0.002183465],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.009353,"threshold_uncertainty_score":0.03572237,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.04522835001101736,"score_gpt":0.3462710113618439,"score_spread":0.3010426613508265,"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."}}