{"id":"W4402670735","doi":"10.18653/v1/2024.findings-acl.566","title":"Bootstrapping LLM-based Task-Oriented Dialogue Agents via Self-Talk","year":2024,"lang":"en","type":"article","venue":"","topic":"Speech and dialogue systems","field":"Computer Science","cited_by":9,"is_retracted":false,"has_abstract":true,"ca_institutions":"Artificial Intelligence in Medicine (Canada)","funders":"","keywords":"Computer science; Bootstrapping (finance); Task (project management); Human–computer interaction; Engineering; Systems engineering; Business","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":"codex-gemma-dda1882f352a","candidate_categories":["insufficient_payload"],"consensus_categories":[],"category_scores_codex":[0.0002945999,0.0001994797,0.0001835095,0.0002194196,0.0001066087,0.0003623619,0.0005947826,0.0000951383,0.00006643005],"category_scores_gemma":[0.00002479736,0.0001695438,0.0001438022,0.0007748164,0.00002402578,0.000421684,0.0001231822,0.000134278,0.0008371312],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.00006273764,"about_ca_system_score_gemma":0.0001696824,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.0002756894,"about_ca_topic_score_gemma":0.00003684308,"domain_scores_codex":[0.9982175,0.00007387203,0.0002960883,0.0005884438,0.0003775399,0.0004465058],"domain_scores_gemma":[0.9990302,0.0001034936,0.00004029478,0.0005699088,0.00005194413,0.00020412],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"theoretical_or_conceptual","study_design_gemma":"simulation_or_modeling","study_design_scores_codex":[0.0001129376,0.003137911,0.006507208,0.002905039,0.002031123,0.006166009,0.01107933,0.004204209,0.06652058,0.4241344,0.3299628,0.1432385],"study_design_scores_gemma":[0.000697062,0.0002208426,0.001487854,0.000126015,0.00002917917,0.00006609079,0.00003818971,0.7727522,0.007703601,0.001766604,0.2145382,0.0005741531],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.006208404,0.0003842027,0.9726163,0.0005595621,0.003706217,0.0003103956,0.000008947663,0.002157065,0.01404897],"genre_scores_gemma":[0.9911503,0.000003123433,0.007243942,0.0006313463,0.0002658324,0.00003248009,0.00002387532,0.00001704338,0.000632033],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.9849419,"threshold_uncertainty_score":0.9999408,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.01466826452131366,"score_gpt":0.2415134788242593,"score_spread":0.2268452143029456,"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."}}