{"id":"W2767260595","doi":"10.1145/3126908.3126964","title":"Understanding error propagation in deep learning neural network (DNN) accelerators and applications","year":2017,"lang":"en","type":"article","venue":"","topic":"Radiation Effects in Electronics","field":"Engineering","cited_by":483,"is_retracted":false,"has_abstract":true,"ca_institutions":"University of British Columbia","funders":"","keywords":"Computer science; Redundancy (engineering); Triple modular redundancy; Modular design; Artificial neural network; Deep learning; Resilience (materials science); Deep neural networks; Software; Computer architecture; Efficient energy use; Computer engineering; Embedded system; Artificial intelligence; Distributed computing; Machine learning; Operating system","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":[],"consensus_categories":[],"category_scores_codex":[0.0001412854,0.00008085215,0.00007915713,0.00004819185,0.0002926556,0.0001526237,0.00009337855,0.00005187477,0.00001061012],"category_scores_gemma":[0.00003031858,0.00008527157,0.00001127209,0.00007822357,0.0000261444,0.0003170172,0.00001893374,0.0001979098,0.000004948209],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0002011759,"about_ca_system_score_gemma":0.000007322948,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.000007570007,"about_ca_topic_score_gemma":0.0002342836,"domain_scores_codex":[0.9994892,0.00001792404,0.0001100748,0.0001128073,0.00006303522,0.0002069789],"domain_scores_gemma":[0.9997101,0.00005564607,0.00003781548,0.0001538472,0.000007670281,0.00003495437],"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":[0.000001380276,0.000002321671,0.007483725,0.00001581607,0.000005831904,4.815141e-7,0.00006682486,0.9741349,0.0001018556,0.01037989,0.00004839251,0.007758566],"study_design_scores_gemma":[0.0001748184,0.00001214485,0.01071058,0.000006320693,0.000003920197,0.000001865355,0.00004976324,0.9866537,0.00007927282,0.001508445,0.0006965109,0.0001026323],"study_design_candidate":"simulation_or_modeling","study_design_consensus":"simulation_or_modeling","genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.4135929,0.0005919644,0.5526419,0.0002561731,0.000299253,0.000971822,4.408855e-7,0.0006312965,0.03101432],"genre_scores_gemma":[0.9993086,0.00003249995,0.0004021748,0.0000153334,0.0001119747,0.00005913761,0.000003080245,0.0000205615,0.00004664435],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.5857157,"threshold_uncertainty_score":0.3477274,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.03428589350407762,"score_gpt":0.2555871987360288,"score_spread":0.2213013052319512,"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."}}