{"id":"W4414870706","doi":"10.1051/0004-6361/202554065","title":"Interpreting deep learning-based stellar mass estimation via causal analysis and mutual information decomposition","year":2025,"lang":"en","type":"article","venue":"Astronomy and Astrophysics","topic":"Astronomical Observations and Instrumentation","field":"Engineering","cited_by":2,"is_retracted":false,"has_abstract":true,"ca_institutions":"","funders":"Lawrence Berkeley National Laboratory; University of California, Los Angeles; National Natural Science Foundation of China; York University; Carnegie Mellon University; Office of Science; Johns Hopkins University; College of Engineering, Michigan State University; Harvard University; Ohio State University; New Mexico State University; University of Portsmouth; Yale University; Vanderbilt University; National Science Foundation; University of Washington; Alfred P. Sloan Foundation; Brookhaven National Laboratory; U.S. Department of Energy; California Institute of Technology; National Aeronautics and Space Administration; Jet Propulsion Laboratory; Princeton University","keywords":"Interpretability; Photometry (optics); Galaxy; Stellar mass; Sky; Mutual information; Concatenation (mathematics); Astrometry","routes":{"ca_aff":false,"ca_fund":true,"ca_venue":false,"about_ca":false,"invisible_to_affiliation_only":true},"retraction":null,"screen":null,"direct_labels":[],"prediction":{"model_version":"codex-gemma-dda1882f352a","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.00004435895,0.0001158175,0.0001320208,0.0001473377,0.0001188708,0.0001033761,0.00003360105,0.00003706232,0.000008555493],"category_scores_gemma":[0.000003404925,0.0001321034,0.0000386056,0.0002221732,0.00002910899,0.0005984125,0.00001624932,0.0001162108,0.000004508334],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.00005306262,"about_ca_system_score_gemma":0.000008181913,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.00001358632,"about_ca_topic_score_gemma":0.000002516272,"domain_scores_codex":[0.9994938,0.00001953482,0.0002178683,0.00009720306,0.00004980776,0.0001217491],"domain_scores_gemma":[0.9997823,0.00003275502,0.00005634647,0.00006438642,0.00002689457,0.00003731508],"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":[0.000008665355,0.000004564355,0.009000338,0.00001325872,0.00008521415,3.047827e-8,0.00005288225,0.5287042,0.0004871052,0.0001444967,0.000001960496,0.4614973],"study_design_scores_gemma":[0.0002689902,0.00004585102,0.03065085,0.00001616196,0.0001467522,8.685445e-8,0.0001384495,0.9660121,0.002065347,0.00005593272,0.0005056704,0.00009377775],"study_design_candidate":"simulation_or_modeling","study_design_consensus":"simulation_or_modeling","genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.4136307,0.000006940234,0.5861474,0.00001585641,0.00002756176,0.00005426534,0.000003609871,0.00004000724,0.00007366565],"genre_scores_gemma":[0.8694327,0.00000180008,0.1301924,0.0000148663,0.00001392177,0.00001031267,0.000326711,0.000004863311,0.000002424179],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.4614035,"threshold_uncertainty_score":0.5387022,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.001972305892795454,"score_gpt":0.1923322088434321,"score_spread":0.1903599029506366,"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."}}