{"id":"W4376123182","doi":"10.48550/arxiv.2305.05577","title":"FAENet: Frame Averaging Equivariant GNN for Materials Modeling","year":2023,"lang":"en","type":"preprint","venue":"arXiv (Cornell University)","topic":"Machine Learning in Materials Science","field":"Materials Science","cited_by":7,"is_retracted":false,"has_abstract":true,"ca_institutions":"McGill University; Université de Montréal; Mila - Quebec Artificial Intelligence Institute","funders":"","keywords":"Equivariant map; Scalability; Computer science; Invariant (physics); Homogeneous space; Frame (networking); Theoretical computer science; Representation (politics); Simple (philosophy); Algorithm; Artificial intelligence; Distributed computing; Mathematics; Pure mathematics; Geometry","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","insufficient_payload"],"consensus_categories":["insufficient_payload"],"category_scores_codex":[0.002313242,0.0006083102,0.0008154967,0.0003573666,0.0005128316,0.0006766015,0.002185761,0.0005003891,0.0009222802],"category_scores_gemma":[0.000472226,0.0006764479,0.0002360854,0.000343601,0.0001934699,0.0003725419,0.002680091,0.0004427532,0.0009919724],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0002986505,"about_ca_system_score_gemma":0.0002928075,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.001194446,"about_ca_topic_score_gemma":0.00003545049,"domain_scores_codex":[0.9956383,0.0003670805,0.0006126724,0.002110085,0.0002442294,0.001027663],"domain_scores_gemma":[0.9972455,0.0002904328,0.0005729443,0.001396516,0.0002586425,0.0002359983],"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":[0.00009517206,0.00002681948,0.00008349025,0.0003026579,0.00002142214,0.00007386834,0.0001900383,0.7980458,0.1878415,0.01315425,0.000152817,0.0000121973],"study_design_scores_gemma":[0.0004972861,0.00005838904,0.00007063241,0.0003245228,0.0001121912,0.000005508158,0.000101875,0.9070314,0.01840285,0.07236645,0.0001795608,0.0008493852],"study_design_candidate":"simulation_or_modeling","study_design_consensus":"simulation_or_modeling","genre_codex":"empirical","genre_gemma":"empirical","genre_scores_codex":[0.5910038,0.00001297088,0.4036247,0.0001246676,0.003555727,0.0006000531,0.0002959445,0.00061799,0.0001641165],"genre_scores_gemma":[0.9885123,0.00005000891,0.008400823,0.0001237301,0.0005343902,0.00001682052,0.0001182448,0.0001241908,0.002119543],"genre_candidate":"empirical","genre_consensus":"empirical","teacher_disagreement_score":0.3975084,"threshold_uncertainty_score":0.999991,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.153828673311559,"score_gpt":0.2338576904466584,"score_spread":0.08002901713509938,"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."}}