{"id":"W4412888345","doi":"10.18653/v1/2025.findings-acl.576","title":"User Behavior Prediction as a Generic, Robust, Scalable, and Low-Cost Evaluation Strategy for Estimating Generalization in LLMs","year":2025,"lang":"en","type":"article","venue":"","topic":"Data Mining Algorithms and Applications","field":"Computer Science","cited_by":1,"is_retracted":false,"has_abstract":true,"ca_institutions":"Artificial Intelligence in Medicine (Canada)","funders":"","keywords":"Generalization; Scalability; Computer science; Artificial intelligence; Machine learning; Mathematics; Database","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":[0.006361771,0.001757829,0.001604975,0.001866256,0.0005833494,0.00177805,0.001922089,0.001898386,0.001663469],"category_scores_gemma":[0.02732903,0.0006742599,0.001308053,0.001226437,0.0009455159,0.004080641,0.002160335,0.00296337,0.001052667],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.001551979,"about_ca_system_score_gemma":0.001762559,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.01032209,"about_ca_topic_score_gemma":0.01431686,"domain_scores_codex":[0.9966614,0.001362805,0.0002384072,0.0009352922,0.0005660191,0.0002360723],"domain_scores_gemma":[0.9872285,0.007056211,0.0006474757,0.003332374,0.001273488,0.0004618786],"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":[0.001287034,0.0006499665,0.04013439,0.0003660957,0.0006304755,0.0002016496,0.0003050477,0.5693594,0.009043495,0.00574921,0.01735248,0.3549207],"study_design_scores_gemma":[0.00002152604,0.0001155072,0.001930756,0.00001625598,0.00002699944,0.00004433441,0.00003305162,0.9901091,0.001556611,0.005449142,0.0006787354,0.00001795596],"study_design_candidate":"simulation_or_modeling","study_design_consensus":"simulation_or_modeling","genre_codex":"methods","genre_gemma":"methods","genre_scores_codex":[0.2073236,0.002024948,0.7587287,0.001962462,0.0002056903,0.0004572452,0.003970301,0.02182429,0.003502727],"genre_scores_gemma":[0.7966494,0.0002588136,0.1939614,0.0006540424,0.0001295976,0.000373175,0.005855518,0.0004555644,0.001662412],"genre_candidate":"methods","genre_consensus":"methods","teacher_disagreement_score":0.01032209,"threshold_uncertainty_score":0.03364462,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.04778180600852341,"score_gpt":0.3289911020724153,"score_spread":0.2812092960638919,"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."}}