{"id":"W3034552719","doi":"10.18653/v1/2020.acl-main.591","title":"Exploiting Syntactic Structure for Better Language Modeling: A Syntactic Distance Approach","year":2020,"lang":"en","type":"preprint","venue":"","topic":"Topic Modeling","field":"Computer Science","cited_by":15,"is_retracted":false,"has_abstract":true,"ca_institutions":"McGill University; Canadian Institute for Advanced Research; Université de Montréal","funders":"National Natural Science Foundation of China; Westlake University","keywords":"Treebank; Computer science; Perplexity; Parsing; Natural language processing; Artificial intelligence; Language model; Task (project management); Representation (politics); Parse tree; Syntactic structure; Syntax","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":[0.001698406,0.001181217,0.0008836119,0.001508417,0.0006581393,0.001560692,0.001678322,0.001335802,0.002229673],"category_scores_gemma":[0.00492217,0.0005158578,0.001160111,0.001655336,0.0008213831,0.004405997,0.002596364,0.002615513,0.001330518],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.000879763,"about_ca_system_score_gemma":0.001106674,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.001404029,"about_ca_topic_score_gemma":0.00239868,"domain_scores_codex":[0.9989538,0.0004560375,0.00005350864,0.0003007103,0.0001718298,0.00006406627],"domain_scores_gemma":[0.9973868,0.001450781,0.0002306773,0.0005237832,0.0002992158,0.0001087386],"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":[0.0003056567,0.0003441041,0.005690875,0.000316532,0.0003568006,0.0002435191,0.0006475417,0.3314175,0.03118237,0.08388212,0.006599764,0.5390133],"study_design_scores_gemma":[0.00001136303,0.00004425799,0.0005818841,0.00001114605,0.0000365004,0.0000556398,0.00003272611,0.9383583,0.003446467,0.05584911,0.001550339,0.00002230216],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.05261881,0.0004422039,0.9423294,0.0007665807,0.00005178836,0.00004522751,0.0003081045,0.001153136,0.002284752],"genre_scores_gemma":[0.683215,0.0005706561,0.3103503,0.0002880218,0.0001838323,0.0001321012,0.001575667,0.0005355611,0.00314888],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.002229673,"threshold_uncertainty_score":0.008982122,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.05290269024819936,"score_gpt":0.2758101971273488,"score_spread":0.2229075068791495,"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."}}