{"id":"W2949221912","doi":"10.48550/arxiv.1807.05887","title":"Toward Interpretable Deep Reinforcement Learning with Linear Model U-Trees","year":2018,"lang":"en","type":"preprint","venue":"arXiv (Cornell University)","topic":"Machine Learning and Data Classification","field":"Computer Science","cited_by":0,"is_retracted":false,"has_abstract":true,"ca_institutions":"Simon Fraser University","funders":"","keywords":"Reinforcement learning; Computer science; Artificial intelligence; Artificial neural network; Tree (set theory); Function (biology); Machine learning; Feature (linguistics); Decision tree; Deep neural networks; 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"],"consensus_categories":[],"category_scores_codex":[0.0002949078,0.0003200244,0.0002932193,0.0002311431,0.0002209368,0.0001858256,0.001791648,0.0002032473,0.00003940355],"category_scores_gemma":[0.00005413482,0.0003135604,0.0001042403,0.0003581513,0.0001107364,0.0005182113,0.002018878,0.0008494989,0.0001687371],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0001678428,"about_ca_system_score_gemma":0.0002050115,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.0001771218,"about_ca_topic_score_gemma":0.00004813487,"domain_scores_codex":[0.9980553,0.0001269601,0.0002017171,0.001114674,0.0001470357,0.0003542869],"domain_scores_gemma":[0.9979283,0.00005114279,0.0003372975,0.001317475,0.0002086128,0.0001571388],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"simulation_or_modeling","study_design_gemma":"simulation_or_modeling","study_design_scores_codex":[0.00004201947,0.00002676079,0.001482298,0.00004582943,0.00005486298,0.00002817542,0.000560462,0.964176,0.00001113769,0.03287194,0.000110398,0.0005901039],"study_design_scores_gemma":[0.0002863233,0.000144177,0.0001981311,0.0001142136,0.00004710855,0.000003744662,0.00006173047,0.9949574,0.00004340774,0.002721327,0.001038748,0.0003837122],"study_design_candidate":"simulation_or_modeling","study_design_consensus":"simulation_or_modeling","genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.01943748,0.00002311345,0.9695548,0.0001276537,0.0001522544,0.0001815989,0.000001824552,0.0004150367,0.01010623],"genre_scores_gemma":[0.9790334,0.00009473719,0.0151128,0.00007876013,0.00007043569,0.000001775373,0.0000841057,0.00002138575,0.005502587],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.9595959,"threshold_uncertainty_score":0.9999316,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.06759661223877765,"score_gpt":0.2031394876909434,"score_spread":0.1355428754521658,"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."}}