{"id":"W3017779903","doi":"10.18653/v1/2020.findings-emnlp.109","title":"Quantifying the Contextualization of Word Representations with Semantic Class Probing","year":2020,"lang":"en","type":"preprint","venue":"","topic":"Topic Modeling","field":"Computer Science","cited_by":12,"is_retracted":false,"has_abstract":true,"ca_institutions":"Microsoft (Canada)","funders":"","keywords":"Contextualization; Computer science; Natural language processing; Task (project management); Artificial intelligence; Context (archaeology); Inference; Word (group theory); Layer (electronics); Embedding; Class (philosophy); Linguistics","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":[0.0009083318,0.00153477,0.0008462137,0.0007339977,0.0004047207,0.001338374,0.0008440872,0.001175417,0.001876875],"category_scores_gemma":[0.006953574,0.0006628816,0.0009194406,0.0008306139,0.001075703,0.004636707,0.001863683,0.002847053,0.0008699989],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0006381205,"about_ca_system_score_gemma":0.0007713946,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.002879797,"about_ca_topic_score_gemma":0.004992345,"domain_scores_codex":[0.9993597,0.00016283,0.0000322541,0.0003148196,0.00005145822,0.00007896256],"domain_scores_gemma":[0.997961,0.001126025,0.000193755,0.0004903729,0.0001372344,0.00009145305],"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":[0.0007237858,0.0003151589,0.01341012,0.0007099043,0.0004765338,0.0002509993,0.001101144,0.3603076,0.1320076,0.017242,0.006819407,0.4666357],"study_design_scores_gemma":[0.00003117033,0.0001250782,0.003249965,0.00004137819,0.0001241729,0.00006650209,0.0001351105,0.9277416,0.01941899,0.04691334,0.002113787,0.00003889542],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.3415034,0.00234918,0.6459735,0.001066061,0.0001473092,0.00007277448,0.0007545697,0.004977667,0.003155587],"genre_scores_gemma":[0.9213752,0.0006342342,0.07478374,0.0002534297,0.00008475195,0.00009232855,0.001199689,0.0004042034,0.00117243],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.002879797,"threshold_uncertainty_score":0.006278753,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.1208758373613004,"score_gpt":0.3257294327598204,"score_spread":0.20485359539852,"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."}}