{"id":"W4405104912","doi":"10.1007/978-3-031-72890-7_9","title":"SAFT: Towards Out-of-Distribution Generalization in Fine-Tuning","year":2024,"lang":"en","type":"book-chapter","venue":"Lecture notes in computer science","topic":"Image and Signal Denoising Methods","field":"Computer Science","cited_by":1,"is_retracted":false,"has_abstract":false,"ca_institutions":"Université de Montréal; Cisco Systems (Canada)","funders":"","keywords":"Computer science; Generalization; Distribution (mathematics); Algorithm; Artificial intelligence; Mathematics; Mathematical analysis","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"],"consensus_categories":[],"category_scores_codex":[0.001541276,0.0003671013,0.0004689357,0.0008035846,0.00009111053,0.0003955571,0.001743524,0.0002781169,0.00001098313],"category_scores_gemma":[0.0001359166,0.0003430549,0.0001190704,0.0009612053,0.0003051463,0.0004960339,0.0008807379,0.0006325653,0.00002224732],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.00032772,"about_ca_system_score_gemma":0.0004936695,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.00004969981,"about_ca_topic_score_gemma":0.00008066536,"domain_scores_codex":[0.997036,0.00006280019,0.0006040415,0.001071637,0.0007763187,0.00044916],"domain_scores_gemma":[0.9985049,0.0002123577,0.0002031275,0.0007784428,0.0002219999,0.00007919231],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"design_other","study_design_gemma":"simulation_or_modeling","study_design_scores_codex":[0.000008242956,0.00002352395,0.00002770308,0.0001242784,0.000009305225,0.0002018562,0.001546125,0.05377708,0.001704437,0.06398863,0.00005894035,0.8785299],"study_design_scores_gemma":[0.0002279007,0.0001008058,0.00009819132,0.0008088777,0.000008484898,0.00002376794,9.78296e-8,0.7433562,0.006562351,0.2463853,0.001987767,0.0004403098],"study_design_candidate":"design_other","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"methods","genre_scores_codex":[0.00006969224,0.0009042855,0.9920297,0.0004941848,0.003730773,0.00020086,0.000009522658,0.00009864638,0.002462395],"genre_scores_gemma":[0.116111,0.0000889874,0.8804134,0.00072519,0.000950425,0.000009770412,0.00005688229,0.00005626397,0.001588146],"genre_candidate":"methods","genre_consensus":"methods","teacher_disagreement_score":0.8780895,"threshold_uncertainty_score":0.9999021,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.02970034779102624,"score_gpt":0.2917598775713738,"score_spread":0.2620595297803476,"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."}}