{"id":"W4322764463","doi":"10.1038/s41597-023-02015-3","title":"The Three Terms Task - an open benchmark to compare human and artificial semantic representations","year":2023,"lang":"en","type":"article","venue":"Scientific Data","topic":"Advanced Text Analysis Techniques","field":"Computer Science","cited_by":2,"is_retracted":false,"has_abstract":true,"ca_institutions":"Université de Montréal; Institut Universitaire de Gériatrie de Montréal","funders":"","keywords":"Computer science; Semantic memory; Natural language processing; Artificial intelligence; Semantic similarity; Task (project management); Benchmark (surveying); Word (group theory); Similarity (geometry); Associative property; Noun; Representation (politics); Cognition; Linguistics; Psychology; Mathematics","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":["sts","scholarly_communication","open_science"],"consensus_categories":["open_science"],"category_scores_codex":[0.00216942,0.0001207953,0.0001536546,0.000223703,0.002221063,0.004993107,0.009421047,0.00001642988,0.00001355662],"category_scores_gemma":[0.0001796635,0.00009184183,0.00001884111,0.002158782,0.0003111929,0.002275651,0.01025833,0.00009851468,0.0001460558],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.00001737889,"about_ca_system_score_gemma":0.00004476262,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.0002026787,"about_ca_topic_score_gemma":0.009899545,"domain_scores_codex":[0.9975235,0.00008811398,0.0003051647,0.0012753,0.0004510306,0.000356831],"domain_scores_gemma":[0.9938155,0.0001170212,0.00009696557,0.005727247,0.00008674231,0.0001565123],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"not_applicable","study_design_gemma":"theoretical_or_conceptual","study_design_scores_codex":[0.000007401924,0.0001555034,0.004162653,0.00001321451,0.00005563197,0.00004101366,0.001829062,0.0001677357,0.03754588,0.1769958,0.6032326,0.1757935],"study_design_scores_gemma":[0.0002182912,0.00009994993,0.06902393,0.00006365542,0.00005082445,0.00001246607,0.0007002301,0.2627042,0.003257497,0.5252708,0.137862,0.0007361621],"study_design_candidate":"not_applicable","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.1757925,0.0000860263,0.7963852,0.01414136,0.002382418,0.003222255,0.001645191,0.002153977,0.004191012],"genre_scores_gemma":[0.9587647,0.00000442477,0.03598879,0.000136594,0.00008868626,0.00008238699,0.002609245,0.00001826304,0.002306885],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.7829722,"threshold_uncertainty_score":0.9990779,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.1460814197492021,"score_gpt":0.4256313126982627,"score_spread":0.2795498929490606,"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."}}