{"id":"W3160563524","doi":"10.18653/v1/2021.acl-long.114","title":"Reflective Decoding: Beyond Unidirectional Generation with Off-the-Shelf Language Models","year":2021,"lang":"en","type":"preprint","venue":"","topic":"Topic Modeling","field":"Computer Science","cited_by":1,"is_retracted":false,"has_abstract":true,"ca_institutions":"","funders":"Army Research Office; Naval Information Warfare Center Pacific; Natural Sciences and Engineering Research Council of Canada; Defense Advanced Research Projects Agency; Allen Institute for Artificial Intelligence","keywords":"Decoding methods; Computer science; Contextualization; Context (archaeology); Sentence; Task (project management); Artificial intelligence; Reflection (computer programming); Unsupervised learning; Natural language processing; Machine learning; Algorithm","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":[0.003593236,0.001257205,0.001243597,0.0008811833,0.001056206,0.005174278,0.00357105,0.002355431,0.01170286],"category_scores_gemma":[0.02163066,0.001101137,0.001253508,0.001211177,0.001896615,0.01079401,0.006599637,0.002943452,0.009885002],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.000742081,"about_ca_system_score_gemma":0.001781461,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.0022702,"about_ca_topic_score_gemma":0.003439725,"domain_scores_codex":[0.997113,0.001502534,0.0001251095,0.0005635415,0.0004711481,0.0002245546],"domain_scores_gemma":[0.9877316,0.006616262,0.0002226401,0.004225847,0.0009558825,0.0002478094],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"design_other","study_design_gemma":"simulation_or_modeling","study_design_scores_codex":[0.0007026728,0.0002812247,0.001708642,0.0004352507,0.0001528679,0.0005566715,0.001812423,0.07783478,0.009998566,0.4089928,0.02963557,0.4678885],"study_design_scores_gemma":[0.00008099759,0.00005357439,0.00007563968,0.00006016321,0.00005192648,0.0001766902,0.0001461334,0.4411351,0.00719445,0.5349088,0.01607936,0.0000371972],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.007375424,0.0005189785,0.9751702,0.0007116295,0.0002036604,0.00006821157,0.0002573272,0.00608835,0.009606165],"genre_scores_gemma":[0.3943288,0.0009718495,0.5785494,0.0008732781,0.0003999968,0.000306905,0.001968201,0.007040844,0.0155607],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.01170286,"threshold_uncertainty_score":0.03915,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.04713021984422529,"score_gpt":0.2817480599094048,"score_spread":0.2346178400651795,"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."}}