{"id":"W4390632530","doi":"10.48550/arxiv.2401.02297","title":"Are LLMs Robust for Spoken Dialogues?","year":2024,"lang":"en","type":"preprint","venue":"arXiv (Cornell University)","topic":"Topic Modeling","field":"Computer Science","cited_by":0,"is_retracted":false,"has_abstract":true,"ca_institutions":"","funders":"Alliance de recherche numérique du Canada; McGill University","keywords":"Perplexity; Robustness (evolution); Spoken language; Computer science; Task (project management); Natural language processing; Artificial intelligence; Set (abstract data type); Speech recognition; Language model","routes":{"ca_aff":false,"ca_fund":true,"ca_venue":false,"about_ca":false,"invisible_to_affiliation_only":true},"retraction":null,"screen":null,"direct_labels":[],"prediction":{"model_version":"metacan-v3-hybrid-931329e0061c","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.006988002,0.002214747,0.001734263,0.001150692,0.0006131905,0.003158731,0.002028042,0.002244728,0.004512923],"category_scores_gemma":[0.04851823,0.0007266783,0.001139525,0.0006809482,0.001173388,0.00333922,0.00270102,0.002949676,0.007401003],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.000822618,"about_ca_system_score_gemma":0.001361165,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.005256607,"about_ca_topic_score_gemma":0.004988668,"domain_scores_codex":[0.991399,0.004189638,0.0004034394,0.002383636,0.001084522,0.0005396301],"domain_scores_gemma":[0.9814528,0.01235,0.0008008355,0.003330965,0.001673172,0.0003921242],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"design_other","study_design_gemma":"bench_or_experimental","study_design_scores_codex":[0.001551639,0.0002783434,0.008856121,0.001562174,0.0008687541,0.0004806898,0.001599678,0.2273276,0.0567441,0.004009412,0.02410986,0.6726116],"study_design_scores_gemma":[0.0001464892,0.0004166405,0.007596218,0.0002679168,0.0002016987,0.0004678107,0.0007799392,0.9084576,0.04983312,0.01836917,0.01329452,0.0001688159],"study_design_candidate":"bench_or_experimental","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.1602976,0.00580233,0.7681423,0.003460024,0.001233579,0.0003332253,0.004128031,0.04719272,0.009410089],"genre_scores_gemma":[0.8514921,0.0009123391,0.1281707,0.001434857,0.0003334889,0.0004718402,0.006939932,0.003973594,0.006271265],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.006988002,"threshold_uncertainty_score":0.03695655,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.17638339622166,"score_gpt":0.1969008740361831,"score_spread":0.02051747781452307,"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."}}