{"id":"W3126624604","doi":"10.1371/journal.pone.0287736","title":"Using deep LSD to build operators in GANs latent space with meaning in real space","year":2023,"lang":"en","type":"article","venue":"PLoS ONE","topic":"Generative Adversarial Networks and Image Synthesis","field":"Computer Science","cited_by":4,"is_retracted":false,"has_abstract":true,"ca_institutions":"TRIUMF","funders":"National Institute of General Medical Sciences; National Institutes of Health; National Science Foundation","keywords":"MNIST database; Autoencoder; Orthonormal basis; Latent variable; Space (punctuation); Eigenvalues and eigenvectors; Dimensionality reduction; Probabilistic latent semantic analysis; Artificial intelligence; Latent class model; Computer science; Pattern recognition (psychology); Representation (politics); Mathematics; Deep learning; Machine learning; Physics","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":[0.001007155,0.0006606168,0.000558333,0.0004565252,0.0002741829,0.0006788529,0.0006729477,0.000641241,0.002706171],"category_scores_gemma":[0.002216106,0.0003643204,0.0007453898,0.0003166923,0.001392617,0.001187716,0.001581992,0.001780629,0.0006023171],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0007581807,"about_ca_system_score_gemma":0.0007098548,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.001604017,"about_ca_topic_score_gemma":0.002553482,"domain_scores_codex":[0.9996268,0.0001372451,0.00001649347,0.00007276309,0.0001019087,0.00004491598],"domain_scores_gemma":[0.999447,0.0002615749,0.00004889116,0.0001079475,0.00007569213,0.00005890694],"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":[0.0001387534,0.00008918426,0.001369462,0.00009434462,0.00004788509,0.0001734648,0.0002388564,0.6822416,0.01719342,0.2139677,0.004528778,0.07991662],"study_design_scores_gemma":[0.000007758948,0.00002383908,0.00009334392,0.000007696813,0.000003072013,0.00002561131,0.00001558565,0.947311,0.001289758,0.05012411,0.001090913,0.000007200561],"study_design_candidate":"simulation_or_modeling","study_design_consensus":"simulation_or_modeling","genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.01683886,0.00008056199,0.9805504,0.0001579575,0.00003623572,0.00002416099,0.00005621362,0.0004627551,0.00179283],"genre_scores_gemma":[0.584644,0.0002315124,0.4047221,0.0004680263,0.00006783022,0.0002154935,0.0003435533,0.000480126,0.008827527],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.002706171,"threshold_uncertainty_score":0.009053051,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.06443923413368993,"score_gpt":0.2504712104728279,"score_spread":0.186031976339138,"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."}}