{"id":"W4389521016","doi":"10.18653/v1/2023.conll-1.12","title":"On the utility of enhancing BERT syntactic bias with Token Reordering Pretraining","year":2023,"lang":"en","type":"article","venue":"","topic":"Topic Modeling","field":"Computer Science","cited_by":0,"is_retracted":false,"has_abstract":true,"ca_institutions":"Université de Montréal; University of Waterloo","funders":"","keywords":"Security token; Computer science; Artificial intelligence; Natural language processing; Natural (archaeology); Parsing; Natural language; Computer security; History","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":[0.002185681,0.001346496,0.001058436,0.0009275772,0.000811199,0.001326591,0.002158129,0.001561171,0.007687429],"category_scores_gemma":[0.008187461,0.0005926783,0.000468684,0.0009881679,0.0008267128,0.004189732,0.001747358,0.002636477,0.004452662],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0006703828,"about_ca_system_score_gemma":0.001522745,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.01144298,"about_ca_topic_score_gemma":0.02323091,"domain_scores_codex":[0.9991289,0.0003762795,0.00005246883,0.0002102776,0.0001130261,0.0001190398],"domain_scores_gemma":[0.9938601,0.004594915,0.0001270291,0.0005752637,0.0006971401,0.0001455224],"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":[0.001618741,0.0005116028,0.004589772,0.0002442203,0.0001593544,0.0002332481,0.0002760612,0.06530293,0.02082513,0.008581466,0.02634637,0.871311],"study_design_scores_gemma":[0.0002074012,0.0003096346,0.001780839,0.0000812252,0.0001730387,0.0001415197,0.0002243055,0.9435491,0.02085192,0.02622445,0.006407624,0.00004891502],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.2478253,0.007316779,0.6861228,0.005289865,0.00153204,0.0002490788,0.001654019,0.01815774,0.0318523],"genre_scores_gemma":[0.7810657,0.001427978,0.1960148,0.001706011,0.0004971362,0.0001806325,0.003140059,0.001662533,0.01430531],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.01144298,"threshold_uncertainty_score":0.02571696,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.07364391134414623,"score_gpt":0.2643656723281962,"score_spread":0.19072176098405,"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."}}