{"id":"W3033434706","doi":"10.48550/arxiv.2006.03535","title":"CoCon: A Self-Supervised Approach for Controlled Text Generation","year":2020,"lang":"en","type":"preprint","venue":"PolyPublie (École Polytechnique de Montréal)","topic":"Topic Modeling","field":"Computer Science","cited_by":16,"is_retracted":false,"has_abstract":true,"ca_institutions":"Université de Montréal","funders":"","keywords":"Computer science; Natural language processing; Phrase; Artificial intelligence; Transformer; Text generation; Control (management); Language model; Content (measure theory); Mathematics","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":["metaepi_narrow","scholarly_communication"],"consensus_categories":[],"category_scores_codex":[0.001266381,0.000846272,0.001411252,0.000469351,0.0003526767,0.001182115,0.002982787,0.0009934845,0.000008827005],"category_scores_gemma":[0.0003330501,0.0008630229,0.0007652699,0.0004197242,0.00004509889,0.0004519913,0.001997181,0.001141731,0.00001035729],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0006892703,"about_ca_system_score_gemma":0.001076171,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.00127894,"about_ca_topic_score_gemma":0.0001500317,"domain_scores_codex":[0.9948634,0.0003301329,0.001203363,0.001930386,0.0006833234,0.0009894199],"domain_scores_gemma":[0.9956715,0.0001996923,0.0006515506,0.002583866,0.0003825208,0.0005108662],"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":[0.0007534471,0.001741218,0.001301095,0.001774285,0.001576831,0.0001198838,0.005749713,0.2600239,0.02583745,0.551957,0.02322955,0.1259356],"study_design_scores_gemma":[0.003123338,0.0001239121,0.00007569346,0.00004163116,0.0001087908,0.0000294837,0.00002423732,0.9860601,0.001894965,0.006551868,0.001142097,0.000823903],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"methods","genre_scores_codex":[0.002452692,0.001602625,0.9757748,0.009158116,0.0005570521,0.006778696,0.00006627507,0.002880926,0.000728821],"genre_scores_gemma":[0.2132204,0.0001176898,0.773642,0.003776625,0.0009743272,0.007727372,0.0001817858,0.00009863338,0.0002611347],"genre_candidate":"methods","genre_consensus":"methods","teacher_disagreement_score":0.7260361,"threshold_uncertainty_score":0.9998547,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.03053735471752665,"score_gpt":0.2444985638956762,"score_spread":0.2139612091781495,"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."}}