{"id":"W4292937774","doi":"10.17760/d20429166","title":"Improving instruction generation for vision-language navigation by reward designing","year":2021,"lang":"en","type":"dissertation","venue":"","topic":"Multimodal Machine Learning Applications","field":"Computer Science","cited_by":0,"is_retracted":false,"has_abstract":true,"ca_institutions":"Science North","funders":"","keywords":"Computer science; Generator (circuit theory); Function (biology); Artificial intelligence; Language model; Human–computer interaction","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":"metacan-v3-hybrid-931329e0061c","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.001248083,0.0009751623,0.0007282003,0.0005789765,0.0003812323,0.0007071425,0.001465889,0.001023081,0.005852962],"category_scores_gemma":[0.005077878,0.000310924,0.0006068838,0.0005015332,0.0005833204,0.001678187,0.001405911,0.001427662,0.001668902],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.000906828,"about_ca_system_score_gemma":0.001387718,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.003283435,"about_ca_topic_score_gemma":0.004742435,"domain_scores_codex":[0.9994124,0.0001715078,0.0000317257,0.000194566,0.0001179895,0.00007194118],"domain_scores_gemma":[0.9985801,0.0006840788,0.00009073065,0.0002511344,0.0003145603,0.00007953541],"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":[0.0004017304,0.000356693,0.003482059,0.0001610726,0.00005100264,0.0001387856,0.0001664633,0.2655995,0.02422936,0.008468973,0.006653455,0.6902909],"study_design_scores_gemma":[0.00002422076,0.0001134317,0.0002984048,0.000008483043,0.00001364876,0.00003872467,0.00001301621,0.9815997,0.01286975,0.003681266,0.00132772,0.00001147212],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.05732292,0.0002635519,0.9309371,0.0002878587,0.0001262436,0.0001481952,0.000183682,0.0082512,0.0024793],"genre_scores_gemma":[0.62783,0.0001406045,0.363546,0.0002829593,0.00005186776,0.000318569,0.0008267947,0.0007187346,0.006284493],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.005852962,"threshold_uncertainty_score":0.01958013,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.009269301080167717,"score_gpt":0.3029517608700136,"score_spread":0.2936824597898459,"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."}}