{"id":"W4414183195","doi":"10.1038/s42256-025-01109-4","title":"Aligning generalization between humans and machines","year":2025,"lang":"en","type":"article","venue":"Nature Machine Intelligence","topic":"Reinforcement Learning in Robotics","field":"Computer Science","cited_by":11,"is_retracted":false,"has_abstract":false,"ca_institutions":"Vector Institute; University of British Columbia","funders":"Nederlandse Organisatie voor Wetenschappelijk Onderzoek","keywords":"Generalization; Abstraction; Cognition; Generative grammar; Key (lock); Human intelligence","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":[0.0003155791,0.0001914324,0.00018543,0.0002464339,0.0002217618,0.0002413173,0.000862533,0.0001985967,0.0000131826],"category_scores_gemma":[0.0002151401,0.0001714258,0.00004283592,0.0006518156,0.00005681716,0.0002957008,0.0004366908,0.0005909957,0.00001278656],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.00003654698,"about_ca_system_score_gemma":0.00003975874,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.00004864853,"about_ca_topic_score_gemma":0.00001093853,"domain_scores_codex":[0.9987093,0.00007142426,0.0003024177,0.0004191046,0.0002584253,0.0002393577],"domain_scores_gemma":[0.9991044,0.0001825408,0.00009998452,0.0004548346,0.0001011916,0.00005699971],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"theoretical_or_conceptual","study_design_gemma":"simulation_or_modeling","study_design_scores_codex":[0.000004229556,0.00001573165,0.1272286,0.00008746506,0.00007251012,0.000008876946,0.000569366,0.1303697,0.0003576574,0.5805565,0.001057833,0.1596715],"study_design_scores_gemma":[0.0001157877,0.00006360663,0.01699368,0.0001135514,0.000028681,0.00000580783,0.00001115524,0.9450496,0.009386623,0.01663416,0.01123959,0.0003577525],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.001660549,0.00150404,0.9889638,0.001460702,0.0004741261,0.0001455932,0.000001507444,0.0002175787,0.005572068],"genre_scores_gemma":[0.9547895,0.0001106704,0.04220333,0.001119862,0.00009465216,0.0000054696,0.00001736428,0.00001001358,0.001649182],"genre_candidate":"methods","genre_consensus":null,"teacher_disagreement_score":0.9531289,"threshold_uncertainty_score":0.6990543,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.00935696220116273,"score_gpt":0.297988163620978,"score_spread":0.2886312014198152,"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."}}