{"id":"W4385571036","doi":"10.18653/v1/2023.findings-acl.120","title":"Pragmatic Inference with a CLIP Listener for Contrastive Captioning","year":2023,"lang":"en","type":"article","venue":"","topic":"Multimodal Machine Learning Applications","field":"Computer Science","cited_by":3,"is_retracted":false,"has_abstract":true,"ca_institutions":"McGill University","funders":"Samsung","keywords":"Closed captioning; Discriminative model; Computer science; Leverage (statistics); Artificial intelligence; Inference; Fluency; Speech recognition; Natural language processing; Prosody; Hyperparameter; Image (mathematics); Linguistics","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":[],"consensus_categories":[],"category_scores_codex":[0.000211385,0.00008128473,0.00009489774,0.0000760147,0.0001599442,0.0001210225,0.0003370461,0.00002443525,0.00001013785],"category_scores_gemma":[0.0002049557,0.00006073795,0.00002342695,0.0004822593,0.00002800591,0.0001859079,0.00007904183,0.00008213267,0.0001638028],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.00001816018,"about_ca_system_score_gemma":0.00005845793,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.0002128223,"about_ca_topic_score_gemma":0.00004186466,"domain_scores_codex":[0.999272,0.00002661446,0.0001186769,0.0002590852,0.0001248339,0.0001987727],"domain_scores_gemma":[0.9988651,0.0006134703,0.00006139014,0.0002889132,0.0001154388,0.00005569935],"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":[0.00001066283,0.00004849884,0.009930144,0.00004477755,0.00003715661,0.000003322564,0.00176716,0.02814701,0.001048478,0.9259799,0.0006575654,0.03232539],"study_design_scores_gemma":[0.0003570273,0.00007262147,0.04813322,0.00002116643,0.000005052739,0.000004852038,0.00004567005,0.9463313,0.00007667247,0.004200518,0.0006338891,0.0001179857],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.02662808,0.000002345141,0.9659983,0.002305496,0.00003169359,0.0003820307,0.000002509804,0.0006459183,0.004003649],"genre_scores_gemma":[0.8195517,5.826654e-7,0.1790939,0.0001645539,0.00002164144,0.0003601982,0.000009019018,0.000007027413,0.0007913524],"genre_candidate":"methods","genre_consensus":null,"teacher_disagreement_score":0.9217793,"threshold_uncertainty_score":0.2476822,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.01852115849155336,"score_gpt":0.3110642857477515,"score_spread":0.2925431272561981,"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."}}