{"id":"W4401656650","doi":"10.1016/j.engappai.2024.109131","title":"Explainability of deep reinforcement learning algorithms in robotic domains by using Layer-wise Relevance Propagation","year":2024,"lang":"en","type":"article","venue":"Engineering Applications of Artificial Intelligence","topic":"Explainable Artificial Intelligence (XAI)","field":"Computer Science","cited_by":9,"is_retracted":false,"has_abstract":true,"ca_institutions":"Institute of Health Economics; University of Alberta","funders":"","keywords":"Computer science; Reinforcement learning; Relevance (law); Artificial intelligence; Layer (electronics); Algorithm; Machine learning; Materials science; Composite material","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":[0.0008364535,0.0002187751,0.0002894293,0.0004107615,0.00009386099,0.00009476423,0.0007616196,0.00009875418,0.00001253962],"category_scores_gemma":[0.00034057,0.0002478033,0.00009086295,0.002102419,0.0001169671,0.0006440722,0.0001606339,0.000328819,0.00002718655],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0002641405,"about_ca_system_score_gemma":0.0001167699,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.0002415798,"about_ca_topic_score_gemma":0.00002094879,"domain_scores_codex":[0.997564,0.00005303625,0.00105492,0.0005711825,0.0003864074,0.0003704592],"domain_scores_gemma":[0.9984379,0.0004184375,0.0001807362,0.000650772,0.000228258,0.00008384767],"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":[0.0000036969,0.00007325894,0.00002086783,0.0001664267,0.00000918264,0.000002149358,0.0008685615,0.7670337,0.05040273,0.1136266,0.000002214267,0.06779066],"study_design_scores_gemma":[0.000007319826,0.00004450974,0.000006508567,0.00009806495,0.000005881741,0.000002541186,0.0002188309,0.6826106,0.312279,0.004426122,0.0001512911,0.0001493353],"study_design_candidate":"simulation_or_modeling","study_design_consensus":"simulation_or_modeling","genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.01636146,0.0006739351,0.9817412,0.000125497,0.0001606912,0.0006705828,0.000001431518,0.0001963514,0.00006880983],"genre_scores_gemma":[0.937009,0.00006025633,0.06264158,0.000004790092,0.00004108234,0.0001925514,0.000006124593,0.00002256459,0.00002206025],"genre_candidate":"methods","genre_consensus":null,"teacher_disagreement_score":0.9206475,"threshold_uncertainty_score":0.9999974,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.02455562071109785,"score_gpt":0.2863314128269965,"score_spread":0.2617757921158986,"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."}}