{"id":"W3165206559","doi":"10.2196/23099","title":"Predicting Semantic Similarity Between Clinical Sentence Pairs Using Transformer Models: Evaluation and Representational Analysis","year":2021,"lang":"en","type":"article","venue":"JMIR Medical Informatics","topic":"Topic Modeling","field":"Computer Science","cited_by":33,"is_retracted":false,"has_abstract":true,"ca_institutions":"","funders":"","keywords":"Natural language processing; Semantic similarity; Computer science; Artificial intelligence; Sentence; Security token; Similarity (geometry); Transformer","routes":{"ca_aff":false,"ca_fund":false,"ca_venue":true,"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":[0.003031845,0.0001308104,0.0003467976,0.0001299632,0.0001559752,0.0001487266,0.0003639087,0.0001986783,0.00005158246],"category_scores_gemma":[0.0005339015,0.0001213983,0.0001501493,0.0007492787,0.0001014993,0.001250716,0.0001975038,0.0004111124,0.000002470893],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.00004982526,"about_ca_system_score_gemma":0.0005062493,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.00002951158,"about_ca_topic_score_gemma":0.0000225673,"domain_scores_codex":[0.9962274,0.000237307,0.001211576,0.0002678894,0.001791309,0.0002645447],"domain_scores_gemma":[0.9982243,0.0004623767,0.0002169582,0.0004449208,0.0003510031,0.0003004477],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"observational","study_design_gemma":"simulation_or_modeling","study_design_scores_codex":[0.000009015603,0.0002112729,0.6593808,0.0003805433,0.001183041,0.00003981882,0.0221702,0.06745872,0.00001029903,0.004925425,0.0001286882,0.2441022],"study_design_scores_gemma":[0.0004394893,0.00001271927,0.01288986,0.00005782758,0.0002426054,0.00002067588,0.0008233279,0.9825739,0.00001857076,0.00278086,0.00001653197,0.0001236086],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.4157886,0.00003055743,0.5831732,0.0004666966,0.00008215903,0.0001286334,0.00000294136,0.00004523056,0.0002819728],"genre_scores_gemma":[0.8984908,0.00003939559,0.1007895,0.0004975959,0.0001285063,0.00001105587,0.00003307401,0.000004334127,0.000005730762],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.9151152,"threshold_uncertainty_score":0.4950478,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.1501242600776237,"score_gpt":0.413109033859521,"score_spread":0.2629847737818973,"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."}}