{"id":"W4385573948","doi":"10.18653/v1/2022.findings-emnlp.440","title":"WordTies: Measuring Word Associations in Language Models via Constrained Sampling","year":2022,"lang":"en","type":"article","venue":"","topic":"Topic Modeling","field":"Computer Science","cited_by":3,"is_retracted":false,"has_abstract":true,"ca_institutions":"","funders":"Natural Sciences and Engineering Research Council of Canada","keywords":"Computer science; Word (group theory); Natural language processing; Associative property; Artificial intelligence; Similarity (geometry); Proxy (statistics); Word Association; Language model; Property (philosophy); Cosine similarity; Machine learning; Linguistics; Mathematics; Pattern recognition (psychology)","routes":{"ca_aff":false,"ca_fund":true,"ca_venue":false,"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":[0.0005840627,0.00008085508,0.0001237562,0.000155027,0.0001910586,0.00007397648,0.000557686,0.00002246422,0.00008114542],"category_scores_gemma":[0.00003664559,0.00009126153,0.00004142169,0.0004090273,0.000008993636,0.0003257282,0.0004190857,0.0002216631,0.000004526479],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0002222296,"about_ca_system_score_gemma":0.00007795703,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.0003971439,"about_ca_topic_score_gemma":0.0001670474,"domain_scores_codex":[0.9988295,0.00007708737,0.0002364293,0.0002774168,0.000320379,0.0002591904],"domain_scores_gemma":[0.9994348,0.0001208618,0.00005696289,0.00032339,0.00002238784,0.00004155307],"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":[0.000002398414,0.00008020489,0.001493277,0.000007231375,0.0000195752,0.00002369857,0.01818009,0.7655953,0.001235388,0.1482515,0.00004134075,0.06506998],"study_design_scores_gemma":[0.000227763,0.000006212014,0.0002473495,0.000005025197,0.000001367207,0.000005010063,0.001064854,0.9815635,0.00006532112,0.01662027,0.00006190867,0.0001313765],"study_design_candidate":"simulation_or_modeling","study_design_consensus":"simulation_or_modeling","genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.06593977,0.00007246178,0.9256143,0.0005323266,0.0001793736,0.0001098683,0.000002846305,0.0002069445,0.00734213],"genre_scores_gemma":[0.8951712,8.585967e-7,0.1041226,0.0002989213,0.00002343742,0.00003347924,0.000002341301,0.000005950962,0.0003411845],"genre_candidate":"methods","genre_consensus":null,"teacher_disagreement_score":0.8292315,"threshold_uncertainty_score":0.3721537,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.07116683802841292,"score_gpt":0.269134695829356,"score_spread":0.197967857800943,"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."}}