{"id":"W3099434687","doi":"10.18653/v1/2020.conll-1.26","title":"“LazImpa”: Lazy and Impatient neural agents learn to communicate efficiently","year":2020,"lang":"en","type":"preprint","venue":"","topic":"Language and cultural evolution","field":"Social Sciences","cited_by":15,"is_retracted":false,"has_abstract":true,"ca_institutions":"","funders":"Agence Nationale de la Recherche; Canadian Institute for Advanced Research; Microsoft Research","keywords":"Zipf's law; Computer science; Channel (broadcasting); Artificial neural network; Natural (archaeology); Artificial intelligence; Theoretical computer science; Computer network; Mathematics","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":"codex-gemma-dda1882f352a","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.0002657918,0.0001703089,0.0002209708,0.00003678443,0.000430851,0.0002188167,0.0005253008,0.0001709886,0.0003258743],"category_scores_gemma":[0.0001787915,0.0001336886,0.00009079419,0.0001671084,0.0001221112,0.00007312595,0.001361683,0.0004017373,0.0001702074],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.000128993,"about_ca_system_score_gemma":0.00007914938,"about_ca_topic_candidate":true,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.01289031,"about_ca_topic_score_gemma":0.004378285,"domain_scores_codex":[0.9984556,0.0002640037,0.0002077103,0.0003487951,0.0003737859,0.0003500796],"domain_scores_gemma":[0.9991347,0.00003355438,0.00009056058,0.0003003648,0.00008169273,0.0003591465],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"qualitative","study_design_gemma":"not_applicable","study_design_scores_codex":[0.0001923769,0.0004062687,0.009626552,0.0002710343,0.0002330643,0.00007047885,0.7387708,0.003401538,0.001130815,0.01618995,0.1205501,0.109157],"study_design_scores_gemma":[0.002265364,0.001032985,0.09749852,0.0005553552,0.0004048836,0.000006386805,0.1878513,0.01855653,0.0003869566,0.005929301,0.6815768,0.003935688],"study_design_candidate":"qualitative","study_design_consensus":null,"genre_codex":"empirical","genre_gemma":"empirical","genre_scores_codex":[0.9415507,0.0004472285,0.00006094733,0.01359641,0.0004837711,0.0007380611,0.00002015732,0.0002048754,0.04289783],"genre_scores_gemma":[0.9937291,0.0001518068,0.0002308257,0.002181655,0.0001793548,0.0000232776,0.00004815867,0.00001096146,0.003444924],"genre_candidate":"empirical","genre_consensus":"empirical","teacher_disagreement_score":0.5610266,"threshold_uncertainty_score":0.9936829,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.07412707764797923,"score_gpt":0.3617043061144944,"score_spread":0.2875772284665152,"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."}}