{"id":"W4412125611","doi":"10.22541/au.175208933.31059664/v1","title":"Knowledge Distillation for Deep Learning-Based Hydrological Prediction and Forecasting","year":2025,"lang":"en","type":"preprint","venue":"","topic":"Hydrological Forecasting Using AI","field":"Environmental Science","cited_by":0,"is_retracted":false,"has_abstract":true,"ca_institutions":"McGill University; University of Waterloo","funders":"Natural Resources Canada","keywords":"Distillation; Computer science; Artificial intelligence; Deep learning; Machine learning; Chemistry","routes":{"ca_aff":true,"ca_fund":true,"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":[0.0006894271,0.0002891734,0.0003235297,0.00006135935,0.0003274549,0.00007034721,0.0001849305,0.0005139697,0.0003633953],"category_scores_gemma":[0.001175825,0.0002401423,0.0001232185,0.0001314493,0.0002397032,0.00004550964,0.000811093,0.0005465437,0.00002177988],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0002069818,"about_ca_system_score_gemma":0.00001830624,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.00006469493,"about_ca_topic_score_gemma":0.00008490431,"domain_scores_codex":[0.9981087,0.0001373219,0.0003981388,0.0008354115,0.0001745155,0.0003459497],"domain_scores_gemma":[0.9988582,0.0006156551,0.0001907827,0.0002043558,0.00002271811,0.0001082821],"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":[0.00005918693,0.00005893002,0.03633654,0.0001099421,0.00001157734,8.225783e-7,0.00006110665,0.9172177,0.00006312967,0.00008425349,0.000198623,0.04579821],"study_design_scores_gemma":[0.0003163223,0.0002021745,0.01282188,0.00006411036,0.00004654745,0.000002006645,0.000002090112,0.9752559,0.00005415238,0.005039282,0.005971463,0.0002240746],"study_design_candidate":"simulation_or_modeling","study_design_consensus":"simulation_or_modeling","genre_codex":"empirical","genre_gemma":"empirical","genre_scores_codex":[0.723538,0.00008825516,0.2339116,0.0003106883,0.0004344245,0.001019683,0.00004101814,0.0005171552,0.04013925],"genre_scores_gemma":[0.9828714,0.000005435856,0.01472936,0.00009829536,0.0001014473,0.0001431116,0.0001581243,0.00001501287,0.001877796],"genre_candidate":"empirical","genre_consensus":"empirical","teacher_disagreement_score":0.2593335,"threshold_uncertainty_score":0.9792721,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.04191905099354637,"score_gpt":0.2735684809529412,"score_spread":0.2316494299593948,"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."}}