{"meta":{"page":1,"per_page":50,"max_per_page":100,"total":3084,"total_is_capped":false,"direct_labels_cover":10,"predictions_cover":3084,"direct_label_status":"direct model label, unvalidated","prediction_status":"machine_predicted_unvalidated (Codex and Gemma teacher distillation)","score_status":"score_only:v0-immature-baseline (scores rank; they never assert a category)","snapshot":{"source":"OpenAlex, pinned release, all 482 partitions","release":"2026-06-24","frame_built":"2026-07-12","author_layer_release":"2026-06-26"},"query_hash":"faf12f59571d","filters":{"topic":"Natural Language Processing Techniques"}},"results":[{"id":"W2157331557","doi":"10.3115/v1/d14-1179","title":"Learning Phrase Representations using RNN Encoder–Decoder for Statistical Machine Translation","year":2014,"lang":"en","type":"preprint","venue":"","topic":"Natural Language Processing Techniques","field":"Computer Science","cited_by":24447,"is_retracted":false,"has_abstract":true,"routes":{"ca_aff":true,"ca_fund":false,"ca_venue":false,"about_ca":false},"ca_institutions":"Polytechnique Montréal","funders":"","keywords":"Machine translation; Computer science; Phrase; Natural language processing; Artificial intelligence; Encoder; Translation (biology); Statistical learning; Speech recognition","authors":[{"name":"Kyunghyun Cho","is_ca":false},{"name":"Bart van Merriënboer","is_ca":false},{"name":"Çağlar Gülçehre","is_ca":false},{"name":"Dzmitry Bahdanau","is_ca":false},{"name":"Fethi Bougares","is_ca":false},{"name":"Holger Schwenk","is_ca":false},{"name":"Yoshua Bengio","is_ca":true}],"retraction":null,"screen_n_in":null,"score":{"opus":0.04515960229356928,"gpt":0.3691423632671048,"spread":0.3239827609735355,"validation_status":"score_only:v0-immature-baseline"},"prediction":{"model_version":"metacan-v3-hybrid-931329e0061c","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.001586163,0.001918219,0.001371472,0.001249252,0.000685956,0.001384774,0.001606542,0.001828301,0.008033043],"category_scores_gemma":[0.006094155,0.0008081711,0.001117079,0.001566833,0.0005848185,0.002990984,0.001656,0.002506482,0.01161398],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0007224108,"about_ca_system_score_gemma":0.001587722,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.003949023,"about_ca_topic_score_gemma":0.008685879,"domain_scores_codex":[0.9987036,0.0005174335,0.00009456237,0.0003597347,0.0002050716,0.0001196416],"domain_scores_gemma":[0.9974133,0.001404553,0.0001103017,0.0004869018,0.0004990729,0.00008582501],"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":[0.0004962503,0.0002762748,0.000725028,0.000339429,0.0001901286,0.0004258547,0.0001624924,0.06922774,0.01665432,0.01549174,0.02941967,0.866591],"study_design_scores_gemma":[0.0000915686,0.0001244218,0.0002217987,0.00003696067,0.00007210032,0.0001604297,0.00006479194,0.9507257,0.01067937,0.0314944,0.006300459,0.00002798022],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.01602018,0.001908612,0.9611914,0.0005021311,0.0005310296,0.000158845,0.001075828,0.01340967,0.005202297],"genre_scores_gemma":[0.2320935,0.001194619,0.7424739,0.0005691656,0.0006027545,0.0004848948,0.008184108,0.001682197,0.01271482],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.008033043,"threshold_uncertainty_score":0.02687317,"prediction_status":"machine_predicted_unvalidated"},"labels":[],"label_agreement":null},{"id":"W2153579005","doi":"10.48550/arxiv.1310.4546","title":"Distributed Representations of Words and Phrases and their Compositionality","year":2013,"lang":"en","type":"article","venue":"arXiv (Cornell University)","topic":"Natural Language Processing Techniques","field":"Computer Science","cited_by":18086,"is_retracted":false,"has_abstract":true,"routes":{"ca_aff":false,"ca_fund":false,"ca_venue":false,"about_ca":true},"ca_institutions":"","funders":"","keywords":"Computer science; Principle of compositionality; Softmax function; Word (group theory); Natural language processing; Artificial intelligence; Simple (philosophy); Semantics (computer science); Quality (philosophy); Speedup; Linguistics; Artificial neural network","authors":[{"name":"Tomáš Mikolov","is_ca":false},{"name":"Ilya Sutskever","is_ca":false},{"name":"Kai Chen","is_ca":false},{"name":"Greg S. Corrado","is_ca":false},{"name":"Jeff Dean","is_ca":false}],"retraction":null,"screen_n_in":null,"score":{"opus":0.03203702315028123,"gpt":0.1988674381578246,"spread":0.1668304150075434,"validation_status":"score_only:v0-immature-baseline"},"prediction":{"model_version":"metacan-v3-hybrid-931329e0061c","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.001217454,0.0006661215,0.0007507542,0.001415567,0.0005151969,0.001154381,0.001078564,0.0008912294,0.003204591],"category_scores_gemma":[0.008046692,0.0005523507,0.0007432161,0.001530511,0.001039984,0.003968123,0.00140981,0.001401382,0.001161066],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0004961253,"about_ca_system_score_gemma":0.0006627681,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.001728678,"about_ca_topic_score_gemma":0.002961219,"domain_scores_codex":[0.9989674,0.0003990796,0.00005032069,0.0002787843,0.0002269606,0.00007735648],"domain_scores_gemma":[0.9978423,0.001232943,0.0002192653,0.0003799872,0.0002614108,0.00006410707],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"design_other","study_design_gemma":"theoretical_or_conceptual","study_design_scores_codex":[0.0004450993,0.0002072804,0.004032312,0.0004273972,0.0001420099,0.0003986181,0.0008155233,0.1971215,0.02848756,0.2064923,0.004316067,0.5571144],"study_design_scores_gemma":[0.00002141123,0.00009001065,0.0007546201,0.00002185364,0.000023289,0.0001164316,0.00007532917,0.8266588,0.002982653,0.1669433,0.002292724,0.00001950582],"study_design_candidate":"theoretical_or_conceptual","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.02890214,0.0002534036,0.9687665,0.0001848934,0.0000267497,0.00004158283,0.0001995576,0.0005430061,0.001082238],"genre_scores_gemma":[0.6138421,0.0006620522,0.3771847,0.0002014662,0.000149935,0.0003088529,0.001418636,0.000246686,0.005985589],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.003204591,"threshold_uncertainty_score":0.01072037,"prediction_status":"machine_predicted_unvalidated"},"labels":[],"label_agreement":null},{"id":"W2964199361","doi":"10.3115/v1/w14-4012","title":"On the Properties of Neural Machine Translation: Encoder–Decoder Approaches","year":2014,"lang":"en","type":"preprint","venue":"","topic":"Natural Language Processing Techniques","field":"Computer Science","cited_by":6616,"is_retracted":false,"has_abstract":true,"routes":{"ca_aff":true,"ca_fund":true,"ca_venue":false,"about_ca":false},"ca_institutions":"Université de Montréal; Canadian Institute for Advanced Research","funders":"Compute Canada; Natural Sciences and Engineering Research Council of Canada; Canada Research Chairs; Canadian Institute for Advanced Research","keywords":"Computer science; Machine translation; Encoder; Translation (biology); Artificial intelligence; Speech recognition; Operating system","authors":[{"name":"Kyunghyun Cho","is_ca":true},{"name":"Bart van Merriënboer","is_ca":true},{"name":"Dzmitry Bahdanau","is_ca":true},{"name":"Yoshua Bengio","is_ca":true}],"retraction":null,"screen_n_in":null,"score":{"opus":0.08784431085000213,"gpt":0.257640388060015,"spread":0.1697960772100129,"validation_status":"score_only:v0-immature-baseline"},"prediction":{"model_version":"metacan-v3-hybrid-931329e0061c","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.006030826,0.001194393,0.001496906,0.001940236,0.0009969762,0.002478506,0.001890862,0.002433633,0.004117437],"category_scores_gemma":[0.03429485,0.001206173,0.00106086,0.002071118,0.003895187,0.009783267,0.002007201,0.004066052,0.001130818],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.002337408,"about_ca_system_score_gemma":0.00154852,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.003465977,"about_ca_topic_score_gemma":0.002454896,"domain_scores_codex":[0.9965501,0.001603697,0.0002436561,0.0007067264,0.000748071,0.00014778],"domain_scores_gemma":[0.9666987,0.02787786,0.001102108,0.00224704,0.001867328,0.0002069838],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"theoretical_or_conceptual","study_design_gemma":"simulation_or_modeling","study_design_scores_codex":[0.0001678802,0.00005544765,0.00104946,0.0003198876,0.00009975925,0.000271403,0.0002360922,0.2275689,0.002859395,0.6900561,0.002033,0.07528273],"study_design_scores_gemma":[0.00001500233,0.00003348304,0.0001920835,0.0000347195,0.00002231915,0.0001045183,0.00001786725,0.6669067,0.001961273,0.3292256,0.00146641,0.00001996407],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.01045157,0.002161613,0.9780821,0.001134237,0.00009474363,0.00004670056,0.0001514331,0.0003033968,0.007574227],"genre_scores_gemma":[0.6080189,0.006369996,0.3739262,0.0009758858,0.0009539538,0.0005574742,0.0008245112,0.0007255271,0.007647597],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.006030826,"threshold_uncertainty_score":0.03189445,"prediction_status":"machine_predicted_unvalidated"},"labels":[],"label_agreement":null},{"id":"W2626778328","doi":"10.65215/2q58a426","title":"Attention Is All You Need","year":2025,"lang":"en","type":"preprint","venue":"","topic":"Natural Language Processing Techniques","field":"Computer Science","cited_by":6569,"is_retracted":false,"has_abstract":true,"routes":{"ca_aff":true,"ca_fund":false,"ca_venue":false,"about_ca":false},"ca_institutions":"University of Toronto","funders":"","keywords":"Computer science; Machine translation; Transformer; BLEU; Encoder; Artificial intelligence; Parallelizable manifold; Parsing; Language model; Natural language processing; Decoding methods; Task (project management); Convolutional neural network; Speech recognition; Machine learning; Algorithm","authors":[{"name":"Ashish Vaswani","is_ca":false},{"name":"Noam Shazeer","is_ca":false},{"name":"Niki Parmar","is_ca":false},{"name":"Jakob Uszkoreit","is_ca":false},{"name":"Llion Jones","is_ca":false},{"name":"Aidan N. Gomez","is_ca":true},{"name":"Łukasz Kaiser","is_ca":false},{"name":"Illia Polosukhin","is_ca":false}],"retraction":null,"screen_n_in":null,"score":{"opus":0.01958296515034699,"gpt":0.3049060813157011,"spread":0.2853231161653542,"validation_status":"score_only:v0-immature-baseline"},"prediction":{"model_version":"metacan-v3-hybrid-931329e0061c","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.001290036,0.000918586,0.0005766302,0.0008549188,0.002028028,0.004122976,0.001106675,0.002498686,0.1262392],"category_scores_gemma":[0.008535541,0.0004529104,0.0006130619,0.0008899282,0.001891009,0.01017581,0.003955811,0.003710305,0.06712462],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.001192543,"about_ca_system_score_gemma":0.001253329,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.004577928,"about_ca_topic_score_gemma":0.006797184,"domain_scores_codex":[0.9988675,0.0002697808,0.00004314049,0.0002760145,0.0003324497,0.00021119],"domain_scores_gemma":[0.9979519,0.0006074531,0.0001224088,0.000474212,0.0004725642,0.0003715249],"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":[0.0001585597,0.00006500304,0.001613597,0.0002840811,0.00004232508,0.0003149215,0.001550167,0.0005070161,0.002492619,0.07725223,0.6155868,0.3001327],"study_design_scores_gemma":[0.00001312493,0.00002882086,0.0009472517,0.0001700805,0.00002948092,0.0003914603,0.000664629,0.0009713371,0.0009812773,0.0611787,0.9345853,0.00003852327],"study_design_candidate":"theoretical_or_conceptual","study_design_consensus":null,"genre_codex":"other","genre_gemma":"methods","genre_scores_codex":[0.01293748,0.009356464,0.1032492,0.1765424,0.009979519,0.0001821989,0.0029446,0.01012203,0.6746862],"genre_scores_gemma":[0.2172754,0.00843675,0.04344456,0.07747998,0.005234098,0.0003113099,0.003860049,0.005544854,0.6384131],"genre_candidate":"methods","genre_consensus":null,"teacher_disagreement_score":0.1262392,"threshold_uncertainty_score":0.4223121,"prediction_status":"machine_predicted_unvalidated"},"labels":[],"label_agreement":null},{"id":"W4243041937","doi":"10.1017/cbo9780511801686","title":"Analyzing Linguistic Data","year":2008,"lang":"en","type":"book","venue":"Cambridge University Press eBooks","topic":"Natural Language Processing Techniques","field":"Computer Science","cited_by":3194,"is_retracted":false,"has_abstract":true,"routes":{"ca_aff":true,"ca_fund":false,"ca_venue":false,"about_ca":false},"ca_institutions":"University of Alberta","funders":"","keywords":"Variety (cybernetics); Computer science; Construct (python library); Variation (astronomy); Visualization; Natural language processing; Range (aeronautics); Artificial intelligence; Measure (data warehouse); Statistical model; Corpus linguistics; Linguistics; Data science; Data mining; Engineering; Programming language","authors":[{"name":"R. Harald Baayen","is_ca":true}],"retraction":null,"screen_n_in":null,"score":{"opus":0.03352730678943484,"gpt":0.2424501016676691,"spread":0.2089227948782343,"validation_status":"score_only:v0-immature-baseline"},"prediction":{"model_version":"metacan-v3-hybrid-931329e0061c","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.005411941,0.000937312,0.0008152378,0.005406387,0.001433904,0.007209702,0.001476171,0.0009199222,0.05099769],"category_scores_gemma":[0.02692075,0.0004977309,0.0008235945,0.005818343,0.001672746,0.00523484,0.002571823,0.002213044,0.03767369],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.001400145,"about_ca_system_score_gemma":0.002211506,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.001365242,"about_ca_topic_score_gemma":0.00198636,"domain_scores_codex":[0.9949432,0.00173065,0.0004572741,0.0006536764,0.002130448,0.00008487212],"domain_scores_gemma":[0.9877977,0.007480753,0.0004641383,0.001837347,0.002269476,0.0001505522],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"design_other","study_design_gemma":"not_applicable","study_design_scores_codex":[0.00002696401,0.00004157253,0.001648363,0.001235269,0.00006349267,0.0002513525,0.004097162,0.001048478,0.002271621,0.1001514,0.2350129,0.6541514],"study_design_scores_gemma":[0.000007934844,0.00003234542,0.001965866,0.0008052386,0.00002612755,0.0004072351,0.002405562,0.002437725,0.001461121,0.08536851,0.9050379,0.00004441099],"study_design_candidate":"not_applicable","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"methods","genre_scores_codex":[0.01072259,0.007347953,0.6669684,0.01276175,0.002731959,0.001383744,0.03114793,0.01014175,0.256794],"genre_scores_gemma":[0.04971389,0.01052129,0.7470227,0.004213885,0.001356526,0.001858593,0.03546619,0.004877033,0.1449699],"genre_candidate":"methods","genre_consensus":"methods","teacher_disagreement_score":0.05099769,"threshold_uncertainty_score":0.1706043,"prediction_status":"machine_predicted_unvalidated"},"labels":[],"label_agreement":null},{"id":"W1592805114","doi":"","title":"Analyzing linguistic data : a practical introduction to statistics using R","year":2008,"lang":"en","type":"book","venue":"","topic":"Natural Language Processing Techniques","field":"Computer Science","cited_by":2077,"is_retracted":false,"has_abstract":true,"routes":{"ca_aff":true,"ca_fund":false,"ca_venue":false,"about_ca":false},"ca_institutions":"University of Alberta","funders":"","keywords":"Variety (cybernetics); Computer science; Variation (astronomy); Construct (python library); Range (aeronautics); Visualization; Natural language processing; Measure (data warehouse); Linguistics; Artificial intelligence; Statistical model; Computational statistics; Data science; Data mining; Machine learning; Engineering","authors":[{"name":"R. Harald Baayen","is_ca":true}],"retraction":null,"screen_n_in":null,"score":{"opus":0.06438462978172278,"gpt":0.3688184923184362,"spread":0.3044338625367134,"validation_status":"score_only:v0-immature-baseline"},"prediction":{"model_version":"metacan-v3-hybrid-931329e0061c","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.00769965,0.001955887,0.002001739,0.003701233,0.000842249,0.004002429,0.002889315,0.00230962,0.05917167],"category_scores_gemma":[0.03045329,0.002101705,0.001763541,0.005024508,0.001835347,0.004937917,0.001949587,0.005964647,0.08156703],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0009599385,"about_ca_system_score_gemma":0.002127981,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.001309381,"about_ca_topic_score_gemma":0.001857521,"domain_scores_codex":[0.9944729,0.0024065,0.0006055127,0.0005106286,0.001917918,0.00008640739],"domain_scores_gemma":[0.9746906,0.0198489,0.00067102,0.001776948,0.002777156,0.0002354246],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"not_applicable","study_design_gemma":"not_applicable","study_design_scores_codex":[0.00001863094,0.00003959185,0.0002674185,0.001330242,0.00008458409,0.0002669269,0.0004819488,0.002408553,0.001136848,0.05373962,0.6551655,0.28506],"study_design_scores_gemma":[0.00001587805,0.00003246352,0.0006770855,0.0005696905,0.00001673884,0.0005497842,0.0001289484,0.003625225,0.0004775641,0.1050539,0.8887912,0.00006158437],"study_design_candidate":"not_applicable","study_design_consensus":"not_applicable","genre_codex":"methods","genre_gemma":"methods","genre_scores_codex":[0.0003626689,0.01870219,0.9054471,0.01036109,0.002624973,0.0004689365,0.005519016,0.01654414,0.03996995],"genre_scores_gemma":[0.003805503,0.01707313,0.9218673,0.004382249,0.002379028,0.001591782,0.003856228,0.007889781,0.03715499],"genre_candidate":"methods","genre_consensus":"methods","teacher_disagreement_score":0.05917167,"threshold_uncertainty_score":0.197949,"prediction_status":"machine_predicted_unvalidated"},"labels":[],"label_agreement":null},{"id":"W2136930489","doi":"10.1162/coli.2006.32.1.13","title":"Evaluating WordNet-based Measures of Lexical Semantic Relatedness","year":2006,"lang":"en","type":"article","venue":"Computational Linguistics","topic":"Natural Language Processing Techniques","field":"Computer Science","cited_by":1414,"is_retracted":false,"has_abstract":true,"routes":{"ca_aff":true,"ca_fund":true,"ca_venue":false,"about_ca":false},"ca_institutions":"University of Toronto","funders":"University of Toronto; University of Pennsylvania","keywords":"WordNet; Computer science; Semantic similarity; Natural language processing; Spelling; Artificial intelligence; Lexical database; Proxy (statistics); Similarity (geometry); Information retrieval; Measure (data warehouse); Linguistics; Machine learning; Data mining","authors":[{"name":"Alexander Budanitsky","is_ca":true},{"name":"Graeme Hirst","is_ca":true}],"retraction":null,"screen_n_in":null,"score":{"opus":0.0587311279945921,"gpt":0.346960390515805,"spread":0.2882292625212128,"validation_status":"score_only:v0-immature-baseline"},"prediction":{"model_version":"metacan-v3-hybrid-931329e0061c","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.008047068,0.001054583,0.001285083,0.01680863,0.0008150082,0.001834905,0.001153538,0.00157927,0.001562804],"category_scores_gemma":[0.03903022,0.000245351,0.000606706,0.009133915,0.001137412,0.00484967,0.001923252,0.0006586149,0.0006350853],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0008571837,"about_ca_system_score_gemma":0.000754552,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.001847559,"about_ca_topic_score_gemma":0.002994315,"domain_scores_codex":[0.9917888,0.003521932,0.001066689,0.0009157127,0.002435959,0.000270954],"domain_scores_gemma":[0.9540678,0.03282277,0.004065515,0.00266195,0.005328716,0.001053136],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"design_other","study_design_gemma":"bench_or_experimental","study_design_scores_codex":[0.002853965,0.001965601,0.2927179,0.001979236,0.00191266,0.0006222777,0.002096339,0.09737049,0.03108677,0.01492058,0.005496279,0.5469779],"study_design_scores_gemma":[0.0003842216,0.004384023,0.2630754,0.0002006437,0.0007751023,0.001771239,0.002096198,0.6443692,0.04514506,0.0313816,0.006037537,0.0003797344],"study_design_candidate":"bench_or_experimental","study_design_consensus":null,"genre_codex":"empirical","genre_gemma":"empirical","genre_scores_codex":[0.9230577,0.003059896,0.06271193,0.0002371843,0.0001257142,0.0002516559,0.001829505,0.0007901886,0.007936302],"genre_scores_gemma":[0.9472552,0.0006117703,0.04735991,0.00003593523,0.00008944846,0.0001533069,0.003636525,0.00009555333,0.0007623129],"genre_candidate":"empirical","genre_consensus":"empirical","teacher_disagreement_score":0.01680863,"threshold_uncertainty_score":0.04255748,"prediction_status":"machine_predicted_unvalidated"},"labels":[],"label_agreement":null},{"id":"W2140679639","doi":"","title":"A Neural Probabilistic Language Model","year":2000,"lang":"en","type":"article","venue":"","topic":"Natural Language Processing Techniques","field":"Computer Science","cited_by":1162,"is_retracted":false,"has_abstract":true,"routes":{"ca_aff":true,"ca_fund":false,"ca_venue":false,"about_ca":false},"ca_institutions":"Université de Montréal; Computer Research Institute of Montréal","funders":"","keywords":"Trigram; Computer science; Artificial intelligence; Probabilistic logic; Word (group theory); Generalization; Natural language processing; Similarity (geometry); Language model; Curse of dimensionality; Sequence (biology); Representation (politics); Function (biology); Statistical model; Artificial neural network; Sentence; Probability distribution; Mathematics","authors":[{"name":"Yoshua Bengio","is_ca":true},{"name":"Réjean Ducharme","is_ca":true},{"name":"Pascal Vincent","is_ca":true}],"retraction":null,"screen_n_in":null,"score":{"opus":0.01070564996388041,"gpt":0.2618203588921802,"spread":0.2511147089282998,"validation_status":"score_only:v0-immature-baseline"},"prediction":{"model_version":"metacan-v3-hybrid-931329e0061c","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.001124283,0.0006516622,0.00107034,0.001108881,0.0005933407,0.002227017,0.002499574,0.002060217,0.008781947],"category_scores_gemma":[0.004360057,0.0005967916,0.001229375,0.001423154,0.0008396971,0.004089278,0.001313494,0.002360455,0.003448648],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.001122188,"about_ca_system_score_gemma":0.001230631,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.00703231,"about_ca_topic_score_gemma":0.006437566,"domain_scores_codex":[0.9992675,0.0002409915,0.00004140195,0.0002326042,0.0001565223,0.00006095421],"domain_scores_gemma":[0.9988428,0.0007757118,0.00007729698,0.00009279602,0.0001599385,0.00005143759],"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":[0.0002327395,0.0001820386,0.001826149,0.0003432871,0.0002551482,0.0004157167,0.0003079402,0.4658278,0.002785804,0.3095597,0.02162032,0.1966433],"study_design_scores_gemma":[0.00001523984,0.00001874153,0.0001575809,0.00001755945,0.00002077967,0.00009327352,0.00001367649,0.886173,0.0001754794,0.1094494,0.003851627,0.00001372862],"study_design_candidate":"simulation_or_modeling","study_design_consensus":"simulation_or_modeling","genre_codex":"methods","genre_gemma":"methods","genre_scores_codex":[0.01384732,0.002049305,0.9620589,0.003674218,0.000254395,0.0001032918,0.002334688,0.001774378,0.01390349],"genre_scores_gemma":[0.6438712,0.003539429,0.3005213,0.001837246,0.0006382834,0.000757179,0.004811869,0.000439671,0.04358394],"genre_candidate":"methods","genre_consensus":"methods","teacher_disagreement_score":0.008781947,"threshold_uncertainty_score":0.02937853,"prediction_status":"machine_predicted_unvalidated"},"labels":[],"label_agreement":null},{"id":"W4230872509","doi":"10.1016/b0-08-044854-2/05234-2","title":"Association for Computational Linguistics","year":2006,"lang":"en","type":"book-chapter","venue":"Elsevier eBooks","topic":"Natural Language Processing Techniques","field":"Computer Science","cited_by":1154,"is_retracted":false,"has_abstract":false,"routes":{"ca_aff":true,"ca_fund":false,"ca_venue":false,"about_ca":false},"ca_institutions":"University of Toronto","funders":"","keywords":"Association (psychology); Linguistics; Applied linguistics; Computer science; Computational linguistics; Natural language processing; Philosophy; Epistemology","authors":[{"name":"G. Hirst","is_ca":true}],"retraction":null,"screen_n_in":null,"score":{"opus":0.01221444471084589,"gpt":0.2608096693829754,"spread":0.2485952246721295,"validation_status":"score_only:v0-immature-baseline"},"prediction":{"model_version":"metacan-v3-hybrid-931329e0061c","candidate_categories":["insufficient_payload"],"consensus_categories":["insufficient_payload"],"category_scores_codex":[0.002118964,0.001843075,0.00286951,0.005397425,0.001713433,0.00550085,0.001845149,0.002011697,0.3387558],"category_scores_gemma":[0.00991881,0.001502314,0.001112413,0.008931608,0.001856162,0.01215914,0.004249661,0.004130051,0.4252813],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0008518009,"about_ca_system_score_gemma":0.003135448,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.003324491,"about_ca_topic_score_gemma":0.003679744,"domain_scores_codex":[0.9985436,0.0002987566,0.0002140091,0.0004504436,0.0004201784,0.0000729729],"domain_scores_gemma":[0.9949856,0.001493235,0.0002463935,0.001627508,0.001405684,0.0002416562],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"not_applicable","study_design_gemma":"not_applicable","study_design_scores_codex":[0.00001678489,0.00002483557,0.0002417915,0.0005518713,0.0000197119,0.00008270214,0.0002255585,0.0001405577,0.0002959945,0.05712921,0.6940867,0.2471843],"study_design_scores_gemma":[0.000007381017,0.000004331056,0.0001840748,0.0003566253,0.00001699107,0.0001343942,0.000059871,0.0003102726,0.0001147566,0.02611255,0.9726907,0.000008144353],"study_design_candidate":"not_applicable","study_design_consensus":"not_applicable","genre_codex":"other","genre_gemma":"other","genre_scores_codex":[0.001081613,0.0196988,0.09998356,0.01360496,0.01541898,0.0002992775,0.01509083,0.009328762,0.8254933],"genre_scores_gemma":[0.009619866,0.02065908,0.05020132,0.00378208,0.003084664,0.0006778912,0.02283772,0.005694959,0.8834425],"genre_candidate":"other","genre_consensus":"other","teacher_disagreement_score":0.6612442,"threshold_uncertainty_score":0.9431846,"prediction_status":"machine_predicted_unvalidated"},"labels":[],"label_agreement":null},{"id":"W2119168550","doi":"10.1162/0891201042544884","title":"The Alignment Template Approach to Statistical Machine Translation","year":2004,"lang":"en","type":"article","venue":"Computational Linguistics","topic":"Natural Language Processing Techniques","field":"Computer Science","cited_by":933,"is_retracted":false,"has_abstract":true,"routes":{"ca_aff":false,"ca_fund":false,"ca_venue":false,"about_ca":true},"ca_institutions":"","funders":"","keywords":"Computer science; NIST; Machine translation; Evaluation of machine translation; Natural language processing; Example-based machine translation; Artificial intelligence; Phrase; Machine translation software usability; Generalization; Transfer-based machine translation; Translation (biology); Task (project management); Context (archaeology); Feature (linguistics); Rule-based machine translation; Word (group theory); Linguistics","authors":[{"name":"Franz Josef Och","is_ca":false},{"name":"Hermann Ney","is_ca":false}],"retraction":null,"screen_n_in":null,"score":{"opus":0.02186480999394037,"gpt":0.2953628508639949,"spread":0.2734980408700545,"validation_status":"score_only:v0-immature-baseline"},"prediction":{"model_version":"metacan-v3-hybrid-931329e0061c","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.002889534,0.001535862,0.001887775,0.002517394,0.0008623259,0.002795553,0.002571376,0.001907262,0.007118283],"category_scores_gemma":[0.007980834,0.0009584452,0.001889473,0.004376983,0.001277208,0.003065383,0.002092547,0.002932003,0.00947412],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0009297866,"about_ca_system_score_gemma":0.002074727,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.00184641,"about_ca_topic_score_gemma":0.001676509,"domain_scores_codex":[0.9961126,0.001807908,0.0002858526,0.0006366392,0.001036355,0.0001206573],"domain_scores_gemma":[0.996811,0.001354415,0.0002526947,0.0009296159,0.0005798683,0.00007237609],"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":[0.0001032152,0.0001195738,0.0004844826,0.0006870677,0.0003779974,0.0002821159,0.0002048647,0.06246563,0.01354581,0.316671,0.02061995,0.5844383],"study_design_scores_gemma":[0.00004647604,0.0001810668,0.0004733773,0.00008944585,0.0001115153,0.0006899307,0.00005223907,0.4675969,0.01183612,0.4172697,0.1015195,0.000133643],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"methods","genre_scores_codex":[0.0003214897,0.0006344055,0.9952965,0.0001381774,0.0001143384,0.00006476677,0.0001183022,0.001227487,0.002084528],"genre_scores_gemma":[0.03202864,0.002575733,0.9558275,0.0003131784,0.0005394523,0.0006263044,0.001209073,0.0008778851,0.006002198],"genre_candidate":"methods","genre_consensus":"methods","teacher_disagreement_score":0.007118283,"threshold_uncertainty_score":0.02381301,"prediction_status":"machine_predicted_unvalidated"},"labels":[],"label_agreement":null},{"id":"W2100664567","doi":"10.3115/v1/p15-1001","title":"On Using Very Large Target Vocabulary for Neural Machine Translation","year":2015,"lang":"en","type":"article","venue":"","topic":"Natural Language Processing Techniques","field":"Computer Science","cited_by":872,"is_retracted":false,"has_abstract":true,"routes":{"ca_aff":true,"ca_fund":true,"ca_venue":false,"about_ca":false},"ca_institutions":"Université de Montréal","funders":"Samsung; Compute Canada; Natural Sciences and Engineering Research Council of Canada; Canada Research Chairs; Canadian Institute for Advanced Research","keywords":"Machine translation; Computer science; Vocabulary; Artificial intelligence; Natural language processing; Joint (building); Computational linguistics; Artificial neural network; Translation (biology); Volume (thermodynamics); Association (psychology); Speech recognition; Linguistics; Engineering; Philosophy","authors":[{"name":"Sébastien Jean","is_ca":true},{"name":"Kyunghyun Cho","is_ca":true},{"name":"Roland Memisevic","is_ca":true},{"name":"Yoshua Bengio","is_ca":true}],"retraction":null,"screen_n_in":null,"score":{"opus":0.04490222195020935,"gpt":0.312351224200445,"spread":0.2674490022502356,"validation_status":"score_only:v0-immature-baseline"},"prediction":{"model_version":"metacan-v3-hybrid-931329e0061c","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.003357789,0.001423357,0.001927944,0.003078663,0.001615439,0.002442612,0.002458054,0.002208272,0.01161907],"category_scores_gemma":[0.01218917,0.0006996281,0.001082063,0.003772133,0.001067747,0.008043237,0.004609047,0.002233578,0.01235754],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0006161327,"about_ca_system_score_gemma":0.001407983,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.006006271,"about_ca_topic_score_gemma":0.01158081,"domain_scores_codex":[0.9962769,0.001849632,0.0003828319,0.0006921433,0.0005176926,0.0002808275],"domain_scores_gemma":[0.9932014,0.003157453,0.0001695791,0.001922007,0.0014254,0.0001241197],"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":[0.0007984407,0.0002790658,0.0006808732,0.0004699129,0.0001891027,0.0003266245,0.0002991986,0.01970365,0.02048291,0.01729874,0.04165705,0.8978143],"study_design_scores_gemma":[0.0003776927,0.0004448215,0.0009117157,0.0001553388,0.0002299777,0.0005572947,0.0006631027,0.7870588,0.02743982,0.143152,0.03889208,0.0001174417],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.03900882,0.005613179,0.9169099,0.001486488,0.001407467,0.000416482,0.0018595,0.01500016,0.01829795],"genre_scores_gemma":[0.3509439,0.002708737,0.6162224,0.00141867,0.000823477,0.000709094,0.01142039,0.001762772,0.01399056],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.01161907,"threshold_uncertainty_score":0.03886962,"prediction_status":"machine_predicted_unvalidated"},"labels":[],"label_agreement":null},{"id":"W1599016936","doi":"","title":"The Winograd Schema Challenge","year":2011,"lang":"en","type":"article","venue":"","topic":"Natural Language Processing Techniques","field":"Computer Science","cited_by":865,"is_retracted":false,"has_abstract":true,"routes":{"ca_aff":true,"ca_fund":false,"ca_venue":false,"about_ca":false},"ca_institutions":"University of Toronto","funders":"","keywords":"Computer science; Schema (genetic algorithms); Ambiguity; Natural language processing; Artificial intelligence; Turing test; Sentence; Turing; Information retrieval; Programming language","authors":[{"name":"Hector J. Levesque","is_ca":true}],"retraction":null,"screen_n_in":null,"score":{"opus":0.03539908514315045,"gpt":0.2591166325644209,"spread":0.2237175474212704,"validation_status":"score_only:v0-immature-baseline"},"prediction":{"model_version":"metacan-v3-hybrid-931329e0061c","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.005606096,0.0005297365,0.001211139,0.001242146,0.004412415,0.007310462,0.003023573,0.004613525,0.04409001],"category_scores_gemma":[0.02591913,0.000673229,0.001134602,0.00279534,0.006164418,0.027611,0.008611195,0.007086279,0.01043049],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.002788191,"about_ca_system_score_gemma":0.002695765,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.003823795,"about_ca_topic_score_gemma":0.0028321,"domain_scores_codex":[0.9951805,0.001854249,0.0003113649,0.001116589,0.001095272,0.0004420133],"domain_scores_gemma":[0.990988,0.004383037,0.0003171433,0.002640021,0.001067161,0.0006047302],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"theoretical_or_conceptual","study_design_gemma":"simulation_or_modeling","study_design_scores_codex":[0.00004136825,0.00001231424,0.0002304,0.00006706834,0.000007696262,0.00007577547,0.0003174763,0.0003205466,0.0000587406,0.926314,0.04584873,0.02670583],"study_design_scores_gemma":[0.0000222051,0.00000570963,0.00008627852,0.00005741122,0.000005527787,0.0002114279,0.0002757332,0.001231238,0.0002078425,0.8497819,0.1481019,0.00001287442],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"other","genre_gemma":"empirical","genre_scores_codex":[0.03664895,0.01186783,0.2469405,0.2642632,0.005475698,0.0002006964,0.003974223,0.003103255,0.4275257],"genre_scores_gemma":[0.6646176,0.01095007,0.1204658,0.04116027,0.005468423,0.0006727921,0.006977824,0.003256379,0.1464308],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.04409001,"threshold_uncertainty_score":0.1474957,"prediction_status":"machine_predicted_unvalidated"},"labels":[],"label_agreement":null},{"id":"W2131462252","doi":"","title":"A Scalable Hierarchical Distributed Language Model","year":2008,"lang":"en","type":"article","venue":"","topic":"Natural Language Processing Techniques","field":"Computer Science","cited_by":852,"is_retracted":false,"has_abstract":true,"routes":{"ca_aff":true,"ca_fund":false,"ca_venue":false,"about_ca":false},"ca_institutions":"University of Toronto","funders":"","keywords":"Language model; Computer science; Hierarchical database model; Cache language model; Artificial intelligence; Scalability; Probabilistic logic; Simple (philosophy); Feature (linguistics); Tree (set theory); Word (group theory); Binary tree; Machine learning; Natural language processing; Data mining; Algorithm; Natural language; Universal Networking Language; Mathematics","authors":[{"name":"Andriy Mnih","is_ca":true},{"name":"Geoffrey E. Hinton","is_ca":true}],"retraction":null,"screen_n_in":null,"score":{"opus":0.01499336283229919,"gpt":0.2616674500190941,"spread":0.2466740871867949,"validation_status":"score_only:v0-immature-baseline"},"prediction":{"model_version":"metacan-v3-hybrid-931329e0061c","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.0008713042,0.0005712882,0.001005692,0.0004925264,0.0005045228,0.001073054,0.002305801,0.001099581,0.005750282],"category_scores_gemma":[0.002722734,0.0005069076,0.0008783713,0.0009679801,0.0005016223,0.002466597,0.001469889,0.001381802,0.002350551],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.001042287,"about_ca_system_score_gemma":0.001899532,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.01310982,"about_ca_topic_score_gemma":0.01527339,"domain_scores_codex":[0.9994161,0.0001417706,0.0000315949,0.0001894454,0.0001564352,0.00006472189],"domain_scores_gemma":[0.9992425,0.0003343688,0.00004887198,0.0001575045,0.0001717276,0.0000449955],"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":[0.0001629403,0.0001033781,0.0007041278,0.0001188071,0.00008614366,0.0001921441,0.00008350994,0.8154629,0.004819315,0.03659778,0.01282406,0.1288449],"study_design_scores_gemma":[0.000007829938,0.000006465237,0.00003865876,0.000001543174,0.000003982342,0.000009696153,0.000003140833,0.9928834,0.000182343,0.006255966,0.0006039061,0.00000290319],"study_design_candidate":"simulation_or_modeling","study_design_consensus":"simulation_or_modeling","genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.01640158,0.0003712164,0.9723628,0.0006421527,0.0001196887,0.000098506,0.000862209,0.003816047,0.0053259],"genre_scores_gemma":[0.5149992,0.0005761949,0.4649035,0.0005244656,0.0002053614,0.0005436594,0.002698427,0.0004911366,0.01505804],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.01310982,"threshold_uncertainty_score":0.02606708,"prediction_status":"machine_predicted_unvalidated"},"labels":[],"label_agreement":null},{"id":"W36903255","doi":"10.3390/molecules28052010","title":"Hierarchical Probabilistic Neural Network Language Model.","year":2005,"lang":"en","type":"article","venue":"International Conference on Artificial Intelligence and Statistics","topic":"Natural Language Processing Techniques","field":"Computer Science","cited_by":840,"is_retracted":false,"has_abstract":true,"routes":{"ca_aff":true,"ca_fund":false,"ca_venue":false,"about_ca":false},"ca_institutions":"Université de Montréal","funders":"National Center for Advancing Translational Sciences","keywords":"Computer science; Language model; Artificial intelligence; Cluster analysis; Hierarchy; Probabilistic logic; Embedding; Generalization; Artificial neural network; Hierarchical clustering; Hierarchical database model; Natural language processing; Machine learning; Data mining; Mathematics","authors":[{"name":"Fréderic Morin","is_ca":true},{"name":"Yoshua Bengio","is_ca":true}],"retraction":null,"screen_n_in":null,"score":{"opus":0.07106443107823439,"gpt":0.3506428053863268,"spread":0.2795783743080924,"validation_status":"score_only:v0-immature-baseline"},"prediction":{"model_version":"metacan-v3-hybrid-931329e0061c","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.001977554,0.001348436,0.001815766,0.001724129,0.0008701258,0.00187601,0.002538478,0.002007599,0.02025467],"category_scores_gemma":[0.00674375,0.0008179068,0.002109803,0.001751219,0.0008248735,0.002778095,0.001591408,0.002739206,0.007420287],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.001686926,"about_ca_system_score_gemma":0.002167754,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.02821097,"about_ca_topic_score_gemma":0.02564965,"domain_scores_codex":[0.9988282,0.0004326765,0.00007640656,0.000334791,0.0001770589,0.0001507804],"domain_scores_gemma":[0.996842,0.002329556,0.0001748815,0.0001827326,0.0003912897,0.00007940917],"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":[0.0002731805,0.0001970079,0.003107936,0.0003879823,0.0002854107,0.0002804672,0.0001838819,0.6900156,0.0004689054,0.06895947,0.03389309,0.2019471],"study_design_scores_gemma":[0.00001681137,0.00001366085,0.000137465,0.00001696587,0.00001862465,0.00003164914,0.00001338189,0.9617651,0.00007702759,0.03472379,0.003176377,0.000009245978],"study_design_candidate":"simulation_or_modeling","study_design_consensus":"simulation_or_modeling","genre_codex":"methods","genre_gemma":"methods","genre_scores_codex":[0.01492251,0.004231634,0.940626,0.00400142,0.0006463347,0.0003615656,0.01071768,0.006975322,0.01751765],"genre_scores_gemma":[0.5465831,0.003572518,0.3872391,0.001685683,0.0006096321,0.00165151,0.01809134,0.0009763881,0.03959071],"genre_candidate":"methods","genre_consensus":"methods","teacher_disagreement_score":0.02821097,"threshold_uncertainty_score":0.06775862,"prediction_status":"machine_predicted_unvalidated"},"labels":[],"label_agreement":null},{"id":"W2050712820","doi":"10.1145/775047.775138","title":"Discovering word senses from text","year":2002,"lang":"en","type":"article","venue":"","topic":"Natural Language Processing Techniques","field":"Computer Science","cited_by":597,"is_retracted":false,"has_abstract":true,"routes":{"ca_aff":true,"ca_fund":true,"ca_venue":false,"about_ca":false},"ca_institutions":"University of Alberta","funders":"Natural Sciences and Engineering Research Council of Canada","keywords":"Word (group theory); Computer science; Cluster analysis; Precision and recall; Set (abstract data type); Similarity (geometry); Centroid; Artificial intelligence; Natural language processing; Feature (linguistics); Feature vector; Element (criminal law); Space (punctuation); Cluster (spacecraft); Domain (mathematical analysis); Recall; Information retrieval; Mathematics; Linguistics; Image (mathematics)","authors":[{"name":"Patrick Pantel","is_ca":true},{"name":"Dekang Lin","is_ca":true}],"retraction":null,"screen_n_in":null,"score":{"opus":0.01633319923786609,"gpt":0.2386434594540469,"spread":0.2223102602161808,"validation_status":"score_only:v0-immature-baseline"},"prediction":{"model_version":"metacan-v3-hybrid-931329e0061c","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.001752223,0.001219283,0.001170246,0.01320937,0.001387881,0.002262923,0.001090056,0.001110854,0.001652997],"category_scores_gemma":[0.01469443,0.0005952127,0.001394505,0.005855973,0.0009647218,0.004321162,0.002165837,0.0008536202,0.001658181],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0006646673,"about_ca_system_score_gemma":0.001421218,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.001935046,"about_ca_topic_score_gemma":0.004160596,"domain_scores_codex":[0.9962621,0.0008185006,0.0005463578,0.001278603,0.0009269146,0.0001674495],"domain_scores_gemma":[0.9916223,0.004516665,0.0009338977,0.001051967,0.001658318,0.0002167939],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"design_other","study_design_gemma":"bench_or_experimental","study_design_scores_codex":[0.0007029556,0.0003875187,0.05773765,0.00245043,0.000482518,0.002354156,0.004287329,0.01106944,0.06191456,0.014102,0.02143087,0.8230805],"study_design_scores_gemma":[0.0002551924,0.0008032661,0.08306502,0.001238367,0.0009964285,0.01125989,0.01255222,0.433298,0.1533567,0.1327879,0.1698433,0.0005437078],"study_design_candidate":"bench_or_experimental","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.420488,0.003555354,0.5399799,0.000856318,0.0003531371,0.001040307,0.01650346,0.007489074,0.009734431],"genre_scores_gemma":[0.356296,0.001437975,0.6149664,0.00021273,0.0001546019,0.0005830666,0.0235802,0.0005921966,0.002176776],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.01320937,"threshold_uncertainty_score":0.009266794,"prediction_status":"machine_predicted_unvalidated"},"labels":[],"label_agreement":null},{"id":"W2587922133","doi":"10.1111/lang.12224","title":"Corpus Use in Language Learning: A Meta‐Analysis","year":2017,"lang":"en","type":"article","venue":"Language Learning","topic":"Natural Language Processing Techniques","field":"Computer Science","cited_by":551,"is_retracted":false,"has_abstract":true,"routes":{"ca_aff":true,"ca_fund":false,"ca_venue":false,"about_ca":false},"ca_institutions":"Université du Québec à Montréal","funders":"","keywords":"Moderation; Meta-analysis; Open data; Computer science; Transfer of learning; Psychology; Sample (material); Language acquisition; Natural language processing; Open science; Applied linguistics; Artificial intelligence; Mathematics education; Linguistics; World Wide Web; Statistics; Machine learning; Mathematics","authors":[{"name":"Alex Boulton","is_ca":true},{"name":"Tom Cobb","is_ca":true}],"retraction":null,"screen_n_in":null,"score":{"opus":0.03466804115790036,"gpt":0.3161270575087146,"spread":0.2814590163508142,"validation_status":"score_only:v0-immature-baseline"},"prediction":{"model_version":"metacan-v3-hybrid-931329e0061c","candidate_categories":["metaepi_broad"],"consensus_categories":[],"category_scores_codex":[0.08739081,0.00277297,0.0146717,0.01268603,0.001153185,0.006096204,0.003107225,0.002367147,0.004228258],"category_scores_gemma":[0.1970932,0.001755777,0.04187864,0.01036332,0.001639975,0.0048016,0.003171998,0.003139754,0.0003534554],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.00436207,"about_ca_system_score_gemma":0.005245195,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.004375144,"about_ca_topic_score_gemma":0.008155971,"domain_scores_codex":[0.893125,0.07224134,0.01934647,0.006852565,0.007499039,0.0009356124],"domain_scores_gemma":[0.7742379,0.1952324,0.01408155,0.007573242,0.007912148,0.0009627622],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"meta_analysis","study_design_gemma":"meta_analysis","study_design_scores_codex":[0.001321905,0.00003899236,0.004825181,0.1770082,0.8037378,0.0001096212,0.0003614807,0.0004450867,0.0002137061,0.0004277757,0.0005835299,0.0109268],"study_design_scores_gemma":[0.000618671,0.0003563759,0.003641828,0.03372718,0.9570143,0.0000953704,0.0002015,0.0003587135,0.0003809952,0.0009372762,0.002629599,0.00003826022],"study_design_candidate":"meta_analysis","study_design_consensus":"meta_analysis","genre_codex":"review","genre_gemma":"empirical","genre_scores_codex":[0.01869604,0.962939,0.01091013,0.0009752963,0.0006745274,0.002773241,0.001750238,0.0002232763,0.001058375],"genre_scores_gemma":[0.6436886,0.3136698,0.02311864,0.00199764,0.0004075264,0.01403607,0.001934212,0.000286541,0.0008609492],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.9853283,"threshold_uncertainty_score":0.4621723,"prediction_status":"machine_predicted_unvalidated"},"labels":[],"label_agreement":null},{"id":"W1965605789","doi":"10.1145/502512.502559","title":"DIRT @SBT@discovery of inference rules from text","year":2001,"lang":"en","type":"article","venue":"","topic":"Natural Language Processing Techniques","field":"Computer Science","cited_by":548,"is_retracted":false,"has_abstract":true,"routes":{"ca_aff":true,"ca_fund":true,"ca_venue":false,"about_ca":false},"ca_institutions":"University of Alberta","funders":"Natural Sciences and Engineering Research Council of Canada","keywords":"Dirt; Computer science; Inference; Rule of inference; Artificial intelligence; Natural language processing; Information retrieval; Data science; Data mining; Cartography; Geography","authors":[{"name":"Dekang Lin","is_ca":true},{"name":"Patrick Pantel","is_ca":true}],"retraction":null,"screen_n_in":null,"score":{"opus":0.01437500932298097,"gpt":0.2750401360449637,"spread":0.2606651267219827,"validation_status":"score_only:v0-immature-baseline"},"prediction":{"model_version":"metacan-v3-hybrid-931329e0061c","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.004447702,0.001065611,0.001366472,0.005606521,0.001580264,0.003566911,0.003518024,0.002119587,0.007314675],"category_scores_gemma":[0.02961596,0.001008009,0.002055133,0.003486798,0.001817869,0.005588131,0.002483728,0.002886067,0.008062948],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.001113806,"about_ca_system_score_gemma":0.002364257,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.004284431,"about_ca_topic_score_gemma":0.008585417,"domain_scores_codex":[0.9934646,0.001608767,0.0005298593,0.002126344,0.002071321,0.0001990568],"domain_scores_gemma":[0.9758351,0.01636643,0.001183338,0.00412929,0.002101627,0.0003843685],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"design_other","study_design_gemma":"bench_or_experimental","study_design_scores_codex":[0.0002038931,0.0003089711,0.007127674,0.0005940963,0.0002943019,0.0006927142,0.0005344043,0.01489848,0.01141672,0.03711486,0.0368349,0.8899791],"study_design_scores_gemma":[0.0001210881,0.00009646243,0.004032381,0.0001769192,0.0002267622,0.001142784,0.0002445521,0.6611633,0.03505079,0.2141217,0.08350752,0.0001157705],"study_design_candidate":"bench_or_experimental","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"methods","genre_scores_codex":[0.007205754,0.0004358421,0.9784175,0.0008671803,0.0001579967,0.0002175874,0.002239921,0.007722128,0.002736117],"genre_scores_gemma":[0.05901876,0.0004167222,0.9221076,0.0005857376,0.0002118549,0.0003441705,0.01170469,0.0009613621,0.004649173],"genre_candidate":"methods","genre_consensus":"methods","teacher_disagreement_score":0.007314675,"threshold_uncertainty_score":0.02447003,"prediction_status":"machine_predicted_unvalidated"},"labels":[],"label_agreement":null},{"id":"W12836875","doi":"10.63317/3gh6v2hd4z55","title":"From TreeBank to PropBank","year":2002,"lang":"en","type":"article","venue":"","topic":"Natural Language Processing Techniques","field":"Computer Science","cited_by":539,"is_retracted":false,"has_abstract":true,"routes":{"ca_aff":false,"ca_fund":false,"ca_venue":false,"about_ca":true},"ca_institutions":"","funders":"","keywords":"Treebank; Computer science; Natural language processing; Artificial intelligence; Annotation","authors":[{"name":"Paul Kingsbury","is_ca":false},{"name":"Martha Palmer","is_ca":false}],"retraction":null,"screen_n_in":null,"score":{"opus":0.01830638065933082,"gpt":0.252359065373384,"spread":0.2340526847140532,"validation_status":"score_only:v0-immature-baseline"},"prediction":{"model_version":"metacan-v3-hybrid-931329e0061c","candidate_categories":["insufficient_payload"],"consensus_categories":[],"category_scores_codex":[0.0004588611,0.000838658,0.0008439979,0.002591837,0.0007493426,0.002936829,0.00107744,0.0007573557,0.589654],"category_scores_gemma":[0.001709296,0.0007358494,0.0002627937,0.004954773,0.0002896017,0.002747433,0.001694165,0.001172703,0.4851124],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0005855115,"about_ca_system_score_gemma":0.0006869199,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.00385532,"about_ca_topic_score_gemma":0.004996414,"domain_scores_codex":[0.999675,0.00003183394,0.00003357637,0.0001325753,0.00009792153,0.00002904584],"domain_scores_gemma":[0.9992343,0.0001820382,0.00008174527,0.000153984,0.000190644,0.0001574118],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"not_applicable","study_design_gemma":"not_applicable","study_design_scores_codex":[0.00040844,0.00007270627,0.0008428475,0.0006984074,0.00002608425,0.0002169646,0.0002627972,0.0002374594,0.002352383,0.004182485,0.7922772,0.1984223],"study_design_scores_gemma":[0.00003383106,0.00001047434,0.00126105,0.0001085199,0.000005874529,0.00006035699,0.000054897,0.0001483977,0.0003105346,0.001515588,0.9964805,0.00001004654],"study_design_candidate":"not_applicable","study_design_consensus":"not_applicable","genre_codex":"dataset","genre_gemma":"methods","genre_scores_codex":[0.006398774,0.004100794,0.007627193,0.002111916,0.000810364,0.0001652037,0.615924,0.0273756,0.3354861],"genre_scores_gemma":[0.01522594,0.00393777,0.01094738,0.000686973,0.0002055435,0.000240409,0.5238849,0.01439984,0.4304712],"genre_candidate":"methods","genre_consensus":null,"teacher_disagreement_score":0.589654,"threshold_uncertainty_score":0.5853088,"prediction_status":"machine_predicted_unvalidated"},"labels":[],"label_agreement":null},{"id":"W2099779943","doi":"10.3115/1621969.1621986","title":"SemEval-2010 task 8","year":2009,"lang":"en","type":"article","venue":"","topic":"Natural Language Processing Techniques","field":"Computer Science","cited_by":497,"is_retracted":false,"has_abstract":true,"routes":{"ca_aff":true,"ca_fund":false,"ca_venue":false,"about_ca":false},"ca_institutions":"University of Ottawa","funders":"","keywords":"SemEval; Computer science; Task (project management); Testbed; Natural language processing; Artificial intelligence; Information extraction; Natural language; Semantics (computer science); Information retrieval; World Wide Web; Programming language","authors":[{"name":"Iris Hendrickx","is_ca":false},{"name":"Su Nam Kim","is_ca":false},{"name":"Zornitsa Kozareva","is_ca":true},{"name":"Preslav Nakov","is_ca":false},{"name":"Diarmuid Ó Séaghdha","is_ca":false},{"name":"Sebastian Padó","is_ca":false},{"name":"Marco Pennacchiotti","is_ca":false},{"name":"Lorenza Romano","is_ca":false},{"name":"Stan Śzpakowicz","is_ca":false}],"retraction":null,"screen_n_in":null,"score":{"opus":0.009039731999430325,"gpt":0.2638369322279143,"spread":0.254797200228484,"validation_status":"score_only:v0-immature-baseline"},"prediction":{"model_version":"metacan-v3-hybrid-931329e0061c","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.005198946,0.004077274,0.002694404,0.00360242,0.002634084,0.004150004,0.004288187,0.00579035,0.03894611],"category_scores_gemma":[0.01313454,0.0007238045,0.002497713,0.002474467,0.001281862,0.006731471,0.006502693,0.003865096,0.03491705],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.002236745,"about_ca_system_score_gemma":0.004148105,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.008651507,"about_ca_topic_score_gemma":0.0183302,"domain_scores_codex":[0.9941576,0.001845016,0.0005736018,0.001757894,0.001121457,0.0005444985],"domain_scores_gemma":[0.9934604,0.002662063,0.0003310729,0.001799342,0.001163988,0.0005829348],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"not_applicable","study_design_gemma":"not_applicable","study_design_scores_codex":[0.0009239333,0.0007528941,0.002662512,0.002236014,0.0001892467,0.0007040164,0.0003737777,0.002035657,0.004587538,0.006513066,0.8471552,0.1318662],"study_design_scores_gemma":[0.00104636,0.0007268325,0.01202867,0.0007276273,0.0001957352,0.003927256,0.002426211,0.05422312,0.02605409,0.03052935,0.8678287,0.0002860241],"study_design_candidate":"not_applicable","study_design_consensus":"not_applicable","genre_codex":"dataset","genre_gemma":"other","genre_scores_codex":[0.1200891,0.007617997,0.07466604,0.008921039,0.007140889,0.004141478,0.5696771,0.07509587,0.1326504],"genre_scores_gemma":[0.09670084,0.0006602401,0.1012153,0.002068324,0.0004721157,0.002432297,0.763127,0.002673945,0.03064996],"genre_candidate":"other","genre_consensus":null,"teacher_disagreement_score":0.03894611,"threshold_uncertainty_score":0.1302878,"prediction_status":"machine_predicted_unvalidated"},"labels":[],"label_agreement":null},{"id":"W1578310937","doi":"10.1007/978-3-319-10106-4_7","title":"Indeterminate Pronouns: The View from Japanese","year":2017,"lang":"en","type":"book-chapter","venue":"Studies in natural language and linguistic theory","topic":"Natural Language Processing Techniques","field":"Computer Science","cited_by":443,"is_retracted":false,"has_abstract":false,"routes":{"ca_aff":true,"ca_fund":false,"ca_venue":false,"about_ca":false},"ca_institutions":"McGill University","funders":"","keywords":"Indeterminate; Linguistics; Philosophy; Mathematics; Pure mathematics","authors":[{"name":"Angelika Kratzer","is_ca":false},{"name":"Junko Shimoyama","is_ca":true}],"retraction":null,"screen_n_in":null,"score":{"opus":0.01968714651575379,"gpt":0.3135701793963078,"spread":0.2938830328805541,"validation_status":"score_only:v0-immature-baseline"},"prediction":{"model_version":"metacan-v3-hybrid-931329e0061c","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.0009178184,0.0006046483,0.0006256487,0.00230212,0.004188353,0.006218386,0.00119757,0.001893783,0.005737178],"category_scores_gemma":[0.002266331,0.0005162713,0.0004252935,0.003642235,0.008588494,0.01235738,0.001980349,0.00337667,0.00076951],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.001773815,"about_ca_system_score_gemma":0.001686422,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.01617555,"about_ca_topic_score_gemma":0.02227263,"domain_scores_codex":[0.999358,0.0001777185,0.00004741332,0.0002012446,0.0001414235,0.00007424458],"domain_scores_gemma":[0.9990209,0.0004578686,0.00008789094,0.0001281897,0.0002139396,0.00009121883],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"theoretical_or_conceptual","study_design_gemma":"theoretical_or_conceptual","study_design_scores_codex":[0.00002314451,0.000008562449,0.0004542342,0.0001451853,0.000006532525,0.0001582923,0.007210646,0.0000901906,0.000403691,0.9665776,0.008716594,0.01620535],"study_design_scores_gemma":[0.00001408512,0.00002119353,0.003581954,0.0001935479,0.00005509701,0.0005480687,0.006704514,0.001103655,0.0006063815,0.7306748,0.2564575,0.00003921552],"study_design_candidate":"theoretical_or_conceptual","study_design_consensus":"theoretical_or_conceptual","genre_codex":"other","genre_gemma":"empirical","genre_scores_codex":[0.086954,0.1187335,0.06704908,0.05677607,0.003289599,0.00004366073,0.000771463,0.0003514529,0.6660313],"genre_scores_gemma":[0.8353371,0.04386903,0.0244215,0.007229367,0.004078382,0.00007875206,0.0005202948,0.0004563019,0.08400927],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.01617555,"threshold_uncertainty_score":0.03216285,"prediction_status":"machine_predicted_unvalidated"},"labels":[],"label_agreement":null},{"id":"W2760656271","doi":"10.18653/v1/w17-4717","title":"Findings of the 2017 Conference on Machine Translation (WMT17)","year":2017,"lang":"en","type":"article","venue":"","topic":"Natural Language Processing Techniques","field":"Computer Science","cited_by":418,"is_retracted":false,"has_abstract":true,"routes":{"ca_aff":true,"ca_fund":false,"ca_venue":false,"about_ca":false},"ca_institutions":"National Research Council Canada","funders":"European Regional Development Fund; Horizon 2020 Framework Programme; Agence Nationale de la Recherche; Univerzita Karlova v Praze; Trinity College Dublin; Science Foundation Ireland; European Commission","keywords":"Chatterjee; Translation (biology); Machine translation; Art history; Computer science; Art; Artificial intelligence; Chemistry","authors":[{"name":"Ondřej Bojar","is_ca":false},{"name":"Christian Federmann","is_ca":false},{"name":"Yvette Graham","is_ca":false},{"name":"Barry Haddow","is_ca":false},{"name":"Philipp Koehn","is_ca":false},{"name":"Christof Monz","is_ca":true}],"retraction":null,"screen_n_in":null,"score":{"opus":0.04852302411475251,"gpt":0.3154419876781849,"spread":0.2669189635634324,"validation_status":"score_only:v0-immature-baseline"},"prediction":{"model_version":"metacan-v3-hybrid-931329e0061c","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.030872,0.003786767,0.003194849,0.01469636,0.004098598,0.0128099,0.003553762,0.004305138,0.06335443],"category_scores_gemma":[0.04272645,0.0009041323,0.002477681,0.01119526,0.002889212,0.01054184,0.009254429,0.006633366,0.06948832],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.004989407,"about_ca_system_score_gemma":0.01438011,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.01813523,"about_ca_topic_score_gemma":0.02621919,"domain_scores_codex":[0.9782381,0.006434896,0.001598976,0.002060077,0.009632913,0.002035026],"domain_scores_gemma":[0.9385006,0.01107933,0.001567481,0.007313407,0.03304956,0.008489511],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"not_applicable","study_design_gemma":"not_applicable","study_design_scores_codex":[0.0003841686,0.0001708449,0.0008766606,0.0005172191,0.00006388574,0.0001451485,0.0001210968,0.0004868272,0.0008799483,0.002714846,0.9385592,0.0550802],"study_design_scores_gemma":[0.0001484242,0.0001524197,0.003742968,0.0006904938,0.0001246968,0.000273318,0.0005522036,0.002363156,0.006834755,0.008806232,0.9762357,0.00007548397],"study_design_candidate":"not_applicable","study_design_consensus":"not_applicable","genre_codex":"editorial","genre_gemma":"other","genre_scores_codex":[0.02367807,0.08333302,0.05796167,0.2046065,0.255079,0.002390971,0.1988698,0.0130436,0.1610374],"genre_scores_gemma":[0.05277961,0.04301537,0.07751177,0.01922636,0.03631927,0.002679765,0.5380719,0.01099163,0.2194043],"genre_candidate":"other","genre_consensus":null,"teacher_disagreement_score":0.06335443,"threshold_uncertainty_score":0.2119417,"prediction_status":"machine_predicted_unvalidated"},"labels":[],"label_agreement":null},{"id":"W2531207078","doi":"10.1162/tacl_a_00067","title":"Fully Character-Level Neural Machine Translation without Explicit Segmentation","year":2017,"lang":"en","type":"article","venue":"Transactions of the Association for Computational Linguistics","topic":"Natural Language Processing Techniques","field":"Computer Science","cited_by":415,"is_retracted":false,"has_abstract":true,"routes":{"ca_aff":false,"ca_fund":true,"ca_venue":false,"about_ca":false},"ca_institutions":"","funders":"York University; Samsung Advanced Institute of Technology; Samsung; Nvidia","keywords":"Computer science; Machine translation; Character (mathematics); Pooling; Encoder; Convolutional neural network; Artificial intelligence; Translation (biology); Natural language processing; Segmentation; Speech recognition; Task (project management); Language model; Representation (politics)","authors":[{"name":"Jason D. Lee","is_ca":false},{"name":"Kyunghyun Cho","is_ca":false},{"name":"Thomas Hofmann","is_ca":false}],"retraction":null,"screen_n_in":null,"score":{"opus":0.0388812594394734,"gpt":0.3152326302466266,"spread":0.2763513708071532,"validation_status":"score_only:v0-immature-baseline"},"prediction":{"model_version":"metacan-v3-hybrid-931329e0061c","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.0004019013,0.0009828565,0.0006612442,0.0004598483,0.0003852232,0.0008666583,0.001083708,0.0008740707,0.006123744],"category_scores_gemma":[0.002153294,0.0003439613,0.0005900062,0.001021107,0.0003785697,0.001810606,0.0009074261,0.001024423,0.004576148],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0005063899,"about_ca_system_score_gemma":0.00116801,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.003681072,"about_ca_topic_score_gemma":0.007645729,"domain_scores_codex":[0.9995779,0.00008596748,0.00003321502,0.0001669619,0.00008662008,0.00004926735],"domain_scores_gemma":[0.9991353,0.0002345918,0.0000637335,0.0003126864,0.0002257725,0.00002789007],"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":[0.0005324047,0.0003176903,0.001353153,0.000491166,0.0001610785,0.0004859472,0.0001808849,0.1845673,0.09449791,0.0183189,0.02080486,0.6782887],"study_design_scores_gemma":[0.00002489029,0.0001186049,0.0005879584,0.00002126631,0.00003288747,0.000222429,0.00002794045,0.9396688,0.03822535,0.01116515,0.009879828,0.0000249355],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.06779244,0.0007854792,0.9006131,0.0003423279,0.0002369238,0.0001127313,0.001358714,0.01466913,0.014089],"genre_scores_gemma":[0.6053543,0.0005226865,0.3657224,0.0003910598,0.000104567,0.0002041542,0.007608874,0.0009344073,0.01915754],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.006123744,"threshold_uncertainty_score":0.020486,"prediction_status":"machine_predicted_unvalidated"},"labels":[],"label_agreement":null},{"id":"W2144746247","doi":"10.3115/1220355.1220401","title":"Confidence estimation for machine translation","year":2004,"lang":"en","type":"article","venue":"","topic":"Natural Language Processing Techniques","field":"Computer Science","cited_by":356,"is_retracted":false,"has_abstract":true,"routes":{"ca_aff":true,"ca_fund":false,"ca_venue":false,"about_ca":false},"ca_institutions":"Université de Montréal","funders":"","keywords":"NIST; Correctness; Computer science; Machine translation; Natural language processing; Context (archaeology); Artificial intelligence; Translation (biology); Estimation; Evaluation of machine translation; Machine learning; Machine translation software usability; Example-based machine translation; Algorithm; Engineering","authors":[{"name":"John Blatz","is_ca":false},{"name":"Erin Fitzgerald","is_ca":false},{"name":"George Foster","is_ca":false},{"name":"Simona Gandrabur","is_ca":true},{"name":"Cyril Goutte","is_ca":false},{"name":"Alex Kulesza","is_ca":false},{"name":"Alberto Sanchís","is_ca":false},{"name":"Nicola Ueffing","is_ca":false}],"retraction":null,"screen_n_in":null,"score":{"opus":0.01942382530659272,"gpt":0.3005074643097555,"spread":0.2810836390031628,"validation_status":"score_only:v0-immature-baseline"},"prediction":{"model_version":"metacan-v3-hybrid-931329e0061c","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.0170647,0.00157136,0.0027111,0.006036328,0.001360917,0.004188172,0.002940698,0.00282462,0.004034267],"category_scores_gemma":[0.204687,0.001125797,0.001831429,0.005621516,0.003084331,0.007656869,0.003694483,0.004639934,0.001465323],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.00171344,"about_ca_system_score_gemma":0.001180665,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.002006905,"about_ca_topic_score_gemma":0.0007858269,"domain_scores_codex":[0.976604,0.01199164,0.001732993,0.002758735,0.006127637,0.0007848838],"domain_scores_gemma":[0.7571788,0.2105353,0.009310734,0.01116813,0.01081116,0.0009960175],"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":[0.0007985792,0.000125007,0.01388713,0.001195285,0.000549535,0.0004198067,0.0005801775,0.3181071,0.0064823,0.2020416,0.007513291,0.4483002],"study_design_scores_gemma":[0.0000366666,0.0001129098,0.002133299,0.0001289275,0.00007521416,0.000362041,0.00004379052,0.8140833,0.005487905,0.1745016,0.002945393,0.0000889362],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"methods","genre_scores_codex":[0.008382616,0.002572977,0.9865919,0.0003575102,0.00007392007,0.0000378062,0.0001631309,0.000537347,0.001282783],"genre_scores_gemma":[0.5637931,0.002450346,0.4279362,0.0003352048,0.001046551,0.0002709323,0.001936212,0.0006121898,0.001619252],"genre_candidate":"methods","genre_consensus":"methods","teacher_disagreement_score":0.0170647,"threshold_uncertainty_score":0.09024787,"prediction_status":"machine_predicted_unvalidated"},"labels":[],"label_agreement":null},{"id":"W2148437670","doi":"10.18653/v1/2023.acl-long","title":"Proceedings of the 61st Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers)","year":2023,"lang":"en","type":"paratext","venue":"","topic":"Natural Language Processing Techniques","field":"Computer Science","cited_by":354,"is_retracted":false,"has_abstract":true,"routes":{"ca_aff":false,"ca_fund":true,"ca_venue":false,"about_ca":true},"ca_institutions":"","funders":"Yuhan; Atomic Energy of Canada Limited; Strong; Styrelsen för Internationellt Utvecklingssamarbete","keywords":"Volume (thermodynamics); Computational linguistics; Association (psychology); Computer science; Linguistics; Natural language processing; Philosophy; Epistemology","authors":[],"retraction":null,"screen_n_in":null,"score":{"opus":0.01096066040400236,"gpt":0.2699824330741872,"spread":0.2590217726701849,"validation_status":"score_only:v0-immature-baseline"},"prediction":{"model_version":"metacan-v3-hybrid-931329e0061c","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.01348262,0.001221272,0.002959837,0.006347352,0.003050588,0.01380765,0.00296286,0.00372875,0.2823546],"category_scores_gemma":[0.03631999,0.001161274,0.001242633,0.005593422,0.002193909,0.01145045,0.004110031,0.00509961,0.2549989],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.002949964,"about_ca_system_score_gemma":0.005826019,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.003973036,"about_ca_topic_score_gemma":0.005693414,"domain_scores_codex":[0.9941081,0.001849118,0.0008966652,0.001095331,0.00167685,0.0003740142],"domain_scores_gemma":[0.9631881,0.01536989,0.001573477,0.004025263,0.01213831,0.003705022],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"not_applicable","study_design_gemma":"not_applicable","study_design_scores_codex":[0.00004659424,0.00003652028,0.0004200318,0.0003754203,0.00002688382,0.00005394563,0.0001203633,0.00002353934,0.0002784349,0.0009880605,0.9481,0.0495302],"study_design_scores_gemma":[0.00001479144,0.00001544828,0.0009237775,0.0005585575,0.00001869274,0.0001212055,0.0002110529,0.000119818,0.00009661205,0.001756597,0.9961444,0.00001899219],"study_design_candidate":"not_applicable","study_design_consensus":"not_applicable","genre_codex":"editorial","genre_gemma":"other","genre_scores_codex":[0.007413968,0.162805,0.03068904,0.1800059,0.3446141,0.001026124,0.02787642,0.007619131,0.2379504],"genre_scores_gemma":[0.02069371,0.0912893,0.0331985,0.03931727,0.06544188,0.001714496,0.04735066,0.008260616,0.6927336],"genre_candidate":"other","genre_consensus":null,"teacher_disagreement_score":0.2823546,"threshold_uncertainty_score":0.9445702,"prediction_status":"machine_predicted_unvalidated"},"labels":[],"label_agreement":null},{"id":"W2124474181","doi":"10.1016/j.system.2004.04.001","title":"Can learners use concordance feedback for writing errors?","year":2004,"lang":"en","type":"article","venue":"System","topic":"Natural Language Processing Techniques","field":"Computer Science","cited_by":346,"is_retracted":false,"has_abstract":false,"routes":{"ca_aff":true,"ca_fund":false,"ca_venue":false,"about_ca":false},"ca_institutions":"Université du Québec à Montréal","funders":"","keywords":"Concordance; Notice; Sentence; Computer science; Natural language processing; Linguistics; Mathematics education; Psychology","authors":[{"name":"Delian Gaskell","is_ca":true},{"name":"Thomas Cobb","is_ca":true}],"retraction":null,"screen_n_in":null,"score":{"opus":0.02152784306383158,"gpt":0.2726472343495525,"spread":0.2511193912857209,"validation_status":"score_only:v0-immature-baseline"},"prediction":{"model_version":"metacan-v3-hybrid-931329e0061c","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.00978669,0.0009476082,0.0009456281,0.001424677,0.0009303218,0.003843267,0.001919598,0.002916744,0.01346289],"category_scores_gemma":[0.1645112,0.0004843885,0.0003884422,0.0009261562,0.0006040014,0.00602962,0.002017555,0.001664478,0.006876153],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0004377165,"about_ca_system_score_gemma":0.001880889,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.001125523,"about_ca_topic_score_gemma":0.001693495,"domain_scores_codex":[0.9854864,0.006536002,0.001194688,0.00166776,0.004681276,0.0004338659],"domain_scores_gemma":[0.8708996,0.08301423,0.007920008,0.01301217,0.02248569,0.002668389],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"design_other","study_design_gemma":"observational","study_design_scores_codex":[0.00143932,0.001453777,0.05897927,0.0009761578,0.00009016829,0.001676189,0.0140226,0.001495567,0.03272572,0.00335719,0.03428494,0.8494992],"study_design_scores_gemma":[0.001694316,0.007233677,0.1206443,0.002877541,0.001110195,0.01547651,0.04175702,0.1815688,0.3709917,0.05592085,0.1996744,0.001050706],"study_design_candidate":"observational","study_design_consensus":null,"genre_codex":"empirical","genre_gemma":"empirical","genre_scores_codex":[0.7319241,0.0008134908,0.1676722,0.01000393,0.001415485,0.0007997706,0.001223643,0.02554579,0.06060167],"genre_scores_gemma":[0.9036042,0.0003714232,0.07983653,0.0009396658,0.0002090854,0.0002014453,0.0006529943,0.0009916944,0.01319302],"genre_candidate":"empirical","genre_consensus":"empirical","teacher_disagreement_score":0.01346289,"threshold_uncertainty_score":0.05175763,"prediction_status":"machine_predicted_unvalidated"},"labels":[],"label_agreement":null},{"id":"W2159755860","doi":"","title":"Batch Tuning Strategies for Statistical Machine Translation","year":2012,"lang":"en","type":"article","venue":"NPARC","topic":"Natural Language Processing Techniques","field":"Computer Science","cited_by":338,"is_retracted":false,"has_abstract":true,"routes":{"ca_aff":true,"ca_fund":false,"ca_venue":true,"about_ca":false},"ca_institutions":"National Research Council Canada","funders":"Johns Hopkins University","keywords":"Computer science; Machine translation; Machine learning; Feature (linguistics); Artificial intelligence; Variety (cybernetics); Support vector machine; Translation (biology); Simple (philosophy); Sentence; Data mining","authors":[{"name":"Colin Cherry","is_ca":true},{"name":"George Foster","is_ca":true}],"retraction":null,"screen_n_in":null,"score":{"opus":0.02839738380637398,"gpt":0.3112397182538902,"spread":0.2828423344475162,"validation_status":"score_only:v0-immature-baseline"},"prediction":{"model_version":"metacan-v3-hybrid-931329e0061c","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.006563127,0.0008993822,0.001083588,0.001298061,0.0009421756,0.001420404,0.001927871,0.001305421,0.005247932],"category_scores_gemma":[0.01597881,0.000652893,0.0006790788,0.001363982,0.0008336183,0.002730816,0.001127122,0.001676277,0.003346708],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0009782065,"about_ca_system_score_gemma":0.001124462,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.00201522,"about_ca_topic_score_gemma":0.003605023,"domain_scores_codex":[0.9961998,0.00200587,0.0002333151,0.0007710664,0.0006238194,0.0001661347],"domain_scores_gemma":[0.9930962,0.004214146,0.0003953807,0.00154118,0.000622271,0.000130784],"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":[0.00122397,0.0005320092,0.002598931,0.0003396382,0.0002297002,0.0001326941,0.0002294854,0.1906141,0.03368352,0.0203078,0.01542012,0.734688],"study_design_scores_gemma":[0.0001207497,0.0002274606,0.001014942,0.00002748548,0.00003904515,0.00013668,0.00005028829,0.9533692,0.01252225,0.02727629,0.005170324,0.00004526402],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.02981577,0.001973137,0.9547424,0.0002848869,0.0001579894,0.0001570596,0.0002427635,0.008066224,0.004559814],"genre_scores_gemma":[0.4267604,0.0005049597,0.5646689,0.0004040052,0.0002260286,0.0005337071,0.0008882373,0.00157279,0.004440979],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.006563127,"threshold_uncertainty_score":0.03470957,"prediction_status":"machine_predicted_unvalidated"},"labels":[],"label_agreement":null},{"id":"W2888541716","doi":"10.18653/v1/d18-1398","title":"Meta-Learning for Low-Resource Neural Machine Translation","year":2018,"lang":"en","type":"article","venue":"","topic":"Natural Language Processing Techniques","field":"Computer Science","cited_by":319,"is_retracted":false,"has_abstract":true,"routes":{"ca_aff":false,"ca_fund":true,"ca_venue":false,"about_ca":false},"ca_institutions":"","funders":"Tencent; Samsung Advanced Institute of Technology; Canadian Institute for Advanced Research; Samsung; Nvidia","keywords":"Computer science; Machine translation; Artificial intelligence; Romanian; Natural language processing; Transfer of learning; Representation (politics); BLEU; Translation (biology); Resource (disambiguation); Machine learning; Frame (networking); Linguistics","authors":[{"name":"Jiatao Gu","is_ca":false},{"name":"Yong Wang","is_ca":false},{"name":"Yun Chen","is_ca":false},{"name":"Victor O. K. Li","is_ca":false},{"name":"Kyunghyun Cho","is_ca":false}],"retraction":null,"screen_n_in":null,"score":{"opus":0.03958221554592636,"gpt":0.2951378261314581,"spread":0.2555556105855317,"validation_status":"score_only:v0-immature-baseline"},"prediction":{"model_version":"metacan-v3-hybrid-931329e0061c","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.002090234,0.001293904,0.001368596,0.0009583265,0.0005764744,0.001309478,0.002406485,0.001789191,0.003481983],"category_scores_gemma":[0.005122149,0.000620242,0.001215951,0.001436203,0.0006719816,0.002834939,0.00160226,0.00217823,0.001776643],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0008652515,"about_ca_system_score_gemma":0.0009060726,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.001123857,"about_ca_topic_score_gemma":0.002029063,"domain_scores_codex":[0.9989098,0.0004950097,0.00008115457,0.000281973,0.0001496875,0.00008233185],"domain_scores_gemma":[0.9981107,0.001033869,0.0001344796,0.0004620809,0.000203631,0.00005535742],"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":[0.0003584467,0.0002957315,0.001284326,0.0003946593,0.0004016053,0.0002396959,0.0001315654,0.4990323,0.01164756,0.01741122,0.00560353,0.4631994],"study_design_scores_gemma":[0.00002132222,0.0000602223,0.0001080497,0.0000147818,0.00002696073,0.00004356051,0.00001134248,0.9824935,0.002636717,0.0135002,0.00107348,0.000009827055],"study_design_candidate":"simulation_or_modeling","study_design_consensus":"simulation_or_modeling","genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.01747895,0.001255017,0.9759318,0.0003304542,0.000116521,0.00005353082,0.0001260305,0.002934858,0.001772853],"genre_scores_gemma":[0.5940461,0.0006266295,0.3987497,0.000599175,0.0002353413,0.0003759741,0.001027815,0.0004399758,0.003899393],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.003481983,"threshold_uncertainty_score":0.01164842,"prediction_status":"machine_predicted_unvalidated"},"labels":[],"label_agreement":null},{"id":"W4389524555","doi":"10.18653/v1/2023.findings-emnlp.936","title":"RWKV: Reinventing RNNs for the Transformer Era","year":2023,"lang":"en","type":"article","venue":"","topic":"Natural Language Processing Techniques","field":"Computer Science","cited_by":303,"is_retracted":false,"has_abstract":true,"routes":{"ca_aff":true,"ca_fund":false,"ca_venue":false,"about_ca":false},"ca_institutions":"Nexen (Canada)","funders":"","keywords":"STELLA (programming language); Zhàng; Art history; Art; Humanities; Philosophy; History; Archaeology; China","authors":[{"name":"Bo Peng","is_ca":false},{"name":"Eric Alcaide","is_ca":false},{"name":"Quentin Anthony","is_ca":false},{"name":"Alon Albalak","is_ca":false},{"name":"Samuel Arcadinho","is_ca":false},{"name":"Stella Biderman","is_ca":false},{"name":"Huanqi Cao","is_ca":false},{"name":"Xin Cheng","is_ca":false},{"name":"Michael Chung","is_ca":false},{"name":"Leon Derczynski","is_ca":false},{"name":"Xingjian Du","is_ca":false},{"name":"Matteo Grella","is_ca":false},{"name":"Kranthi Kiran GV","is_ca":false},{"name":"Xuzheng He","is_ca":false},{"name":"Haowen Hou","is_ca":false},{"name":"Przemysław Kazienko","is_ca":false},{"name":"Jan Kocoń","is_ca":false},{"name":"Jiaming Kong","is_ca":false},{"name":"Bartłomiej Koptyra","is_ca":false},{"name":"H.C.W. Lau","is_ca":false},{"name":"Jiaju Lin","is_ca":false},{"name":"Krishna Sri Ipsit Mantri","is_ca":false},{"name":"Ferdinand Mom","is_ca":true},{"name":"Atsushi Saito","is_ca":false},{"name":"Guangyu Song","is_ca":false},{"name":"Xiangru Tang","is_ca":false},{"name":"Johan S. Wind","is_ca":false},{"name":"Stanisław Woźniak","is_ca":false},{"name":"Zhenyuan Zhang","is_ca":false},{"name":"Qinghua Zhou","is_ca":false},{"name":"Jian Zhu","is_ca":false},{"name":"Ruijie Zhu","is_ca":false}],"retraction":null,"screen_n_in":null,"score":{"opus":0.02679878535284542,"gpt":0.3004265835493182,"spread":0.2736277981964728,"validation_status":"score_only:v0-immature-baseline"},"prediction":{"model_version":"metacan-v3-hybrid-931329e0061c","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.002719742,0.002035334,0.0009737381,0.001534136,0.0007548918,0.002301592,0.003705812,0.001956325,0.0210024],"category_scores_gemma":[0.009341056,0.001338088,0.00128415,0.0008499018,0.0007677916,0.008092727,0.003730047,0.003643585,0.01828944],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0008235325,"about_ca_system_score_gemma":0.001461307,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.009399713,"about_ca_topic_score_gemma":0.02041159,"domain_scores_codex":[0.9985834,0.0003862891,0.0001075352,0.0004641054,0.0003304528,0.0001282166],"domain_scores_gemma":[0.9976357,0.001128383,0.0000882548,0.0006332645,0.0004192966,0.00009504065],"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":[0.001066195,0.0001990066,0.001170723,0.0006762093,0.0003245737,0.0004271955,0.0003294883,0.04510872,0.01474764,0.01746338,0.1546223,0.7638646],"study_design_scores_gemma":[0.0002408531,0.0001970972,0.0003111641,0.0001536528,0.0001233304,0.0002384144,0.0001138912,0.8455394,0.02231022,0.03894074,0.09174083,0.00009045861],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.008089595,0.001755177,0.7423385,0.0007742451,0.001635239,0.0001469416,0.002678652,0.2345413,0.008040421],"genre_scores_gemma":[0.1876192,0.001865701,0.7183748,0.001720047,0.0006159995,0.0004378697,0.01708749,0.03780759,0.03447128],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.0210024,"threshold_uncertainty_score":0.07026005,"prediction_status":"machine_predicted_unvalidated"},"labels":[],"label_agreement":null},{"id":"W2051593977","doi":"10.1162/coli_a_00002","title":"Generating Phrasal and Sentential Paraphrases: A Survey of Data-Driven Methods","year":2010,"lang":"en","type":"article","venue":"Computational Linguistics","topic":"Natural Language Processing Techniques","field":"Computer Science","cited_by":302,"is_retracted":false,"has_abstract":true,"routes":{"ca_aff":false,"ca_fund":true,"ca_venue":false,"about_ca":false},"ca_institutions":"","funders":"Atomic Energy of Canada Limited; National Science Foundation","keywords":"Paraphrase; Computer science; Natural language processing; Task (project management); Artificial intelligence; Parallel corpora; Field (mathematics); Linguistics; Machine translation","authors":[{"name":"Nitin Madnani","is_ca":false},{"name":"Bonnie J. Dorr","is_ca":false}],"retraction":null,"screen_n_in":null,"score":{"opus":0.06410955906269462,"gpt":0.401568733933962,"spread":0.3374591748712674,"validation_status":"score_only:v0-immature-baseline"},"prediction":{"model_version":"metacan-v3-hybrid-931329e0061c","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.01047882,0.001759898,0.002634373,0.007297795,0.0009583142,0.0034426,0.005062104,0.001920267,0.003921885],"category_scores_gemma":[0.03765097,0.001163737,0.00169058,0.009864944,0.001401756,0.005885391,0.002230366,0.002093506,0.003457798],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0009461999,"about_ca_system_score_gemma":0.001856804,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.001743503,"about_ca_topic_score_gemma":0.002342924,"domain_scores_codex":[0.9888002,0.005245849,0.001435808,0.001457733,0.002913964,0.0001464791],"domain_scores_gemma":[0.9483508,0.03835462,0.001861513,0.004715962,0.006364801,0.0003522425],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"design_other","study_design_gemma":"not_applicable","study_design_scores_codex":[0.0001780643,0.0003150378,0.001234041,0.00376138,0.000109724,0.0001086523,0.0006247784,0.00402025,0.007104069,0.006241766,0.005496593,0.9708057],"study_design_scores_gemma":[0.0006051477,0.001691794,0.01181951,0.00417152,0.00088692,0.006003825,0.004444236,0.4604189,0.1392234,0.1208381,0.2492936,0.0006031058],"study_design_candidate":"not_applicable","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"review","genre_scores_codex":[0.01209397,0.0309357,0.943764,0.001370584,0.0001220495,0.0009578547,0.002346252,0.004271593,0.004137907],"genre_scores_gemma":[0.05460917,0.01910826,0.9149152,0.0004581482,0.000162799,0.0008902143,0.007204855,0.0008128833,0.001838503],"genre_candidate":"review","genre_consensus":null,"teacher_disagreement_score":0.01047882,"threshold_uncertainty_score":0.05541795,"prediction_status":"machine_predicted_unvalidated"},"labels":[],"label_agreement":null},{"id":"W1485211753","doi":"","title":"Continuum companion to systemic functional linguistics","year":2009,"lang":"en","type":"book","venue":"","topic":"Natural Language Processing Techniques","field":"Computer Science","cited_by":281,"is_retracted":false,"has_abstract":true,"routes":{"ca_aff":false,"ca_fund":false,"ca_venue":false,"about_ca":true},"ca_institutions":"","funders":"","keywords":"Context (archaeology); Index (typography); Sociology; Applied linguistics; Library science; Linguistics; Art history; History; Philosophy; Archaeology; Computer science","authors":[{"name":"Μ. Α. Κ. Halliday","is_ca":false},{"name":"Jonathan J. Webster","is_ca":false}],"retraction":null,"screen_n_in":null,"score":{"opus":0.01478926830740126,"gpt":0.2523402590825736,"spread":0.2375509907751723,"validation_status":"score_only:v0-immature-baseline"},"prediction":{"model_version":"metacan-v3-hybrid-931329e0061c","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.0011318,0.001070182,0.00107564,0.003739267,0.001855859,0.004446526,0.0009801679,0.001531971,0.1534741],"category_scores_gemma":[0.003385578,0.0005665084,0.0007281118,0.004923536,0.00210835,0.005222548,0.001890256,0.002741358,0.09172797],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.003378343,"about_ca_system_score_gemma":0.002598209,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.003996243,"about_ca_topic_score_gemma":0.005784364,"domain_scores_codex":[0.9992535,0.0002413011,0.00008181913,0.0001500568,0.0002238907,0.00004941565],"domain_scores_gemma":[0.9982614,0.0008984642,0.00005079667,0.0001932765,0.0004722272,0.0001238348],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"not_applicable","study_design_gemma":"not_applicable","study_design_scores_codex":[0.000008826185,0.00001198617,0.00007087864,0.0002670884,0.000004684723,0.000058982,0.0002769845,0.0001159975,0.0001292328,0.07108809,0.8479085,0.08005889],"study_design_scores_gemma":[0.000002084549,0.000002726516,0.0000968936,0.0001306841,0.000001230404,0.00008459758,0.00005577727,0.00008118463,0.00002625351,0.01534982,0.9841657,0.000003163432],"study_design_candidate":"not_applicable","study_design_consensus":"not_applicable","genre_codex":"other","genre_gemma":"other","genre_scores_codex":[0.0005844922,0.08788698,0.05642905,0.02199957,0.02072284,0.0001598597,0.003752919,0.002621012,0.8058433],"genre_scores_gemma":[0.01594546,0.05916836,0.02354996,0.005529554,0.01126998,0.0003931197,0.005848928,0.002003742,0.8762909],"genre_candidate":"other","genre_consensus":"other","teacher_disagreement_score":0.1534741,"threshold_uncertainty_score":0.5134219,"prediction_status":"machine_predicted_unvalidated"},"labels":[],"label_agreement":null},{"id":"W1978620866","doi":"10.1162/coli.2008.34.2.145","title":"Semantic Role Labeling: An Introduction to the Special Issue","year":2008,"lang":"en","type":"article","venue":"Computational Linguistics","topic":"Natural Language Processing Techniques","field":"Computer Science","cited_by":278,"is_retracted":false,"has_abstract":true,"routes":{"ca_aff":true,"ca_fund":true,"ca_venue":false,"about_ca":false},"ca_institutions":"Canada Research Chairs; Artificial Intelligence in Medicine (Canada); University of Toronto","funders":"Natural Sciences and Engineering Research Council of Canada","keywords":"Computer science; Task (project management); Field (mathematics); Computational linguistics; Identification (biology); Computational model; Semantic role labeling; Artificial intelligence; Natural language processing; Data science; Key (lock); Semantics (computer science)","authors":[{"name":"Lluı́s Màrquez","is_ca":true},{"name":"Xavier Carreras","is_ca":true},{"name":"Kenneth C. Litkowski","is_ca":true},{"name":"Suzanne Stevenson","is_ca":true}],"retraction":null,"screen_n_in":null,"score":{"opus":0.01392567728635698,"gpt":0.2788210233112759,"spread":0.2648953460249189,"validation_status":"score_only:v0-immature-baseline"},"prediction":{"model_version":"metacan-v3-hybrid-931329e0061c","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.003817555,0.001616728,0.00183728,0.004672071,0.00277345,0.007790515,0.001989048,0.004152158,0.03282798],"category_scores_gemma":[0.01049643,0.001086494,0.001887413,0.003595748,0.00312039,0.01329346,0.00341756,0.01004796,0.02249422],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.002023924,"about_ca_system_score_gemma":0.002021655,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.00105027,"about_ca_topic_score_gemma":0.001758201,"domain_scores_codex":[0.9981395,0.0005108117,0.0002764369,0.0004275127,0.0005210843,0.0001247214],"domain_scores_gemma":[0.9905044,0.005801042,0.0004638284,0.0006946858,0.001738897,0.0007971383],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"not_applicable","study_design_gemma":"not_applicable","study_design_scores_codex":[0.00002713899,0.00006760091,0.0001974107,0.0003108014,0.00001582576,0.00009813481,0.0001373242,0.0002542073,0.0004874906,0.02225694,0.8864222,0.08972489],"study_design_scores_gemma":[0.000003936943,0.00002468284,0.0002786157,0.0002618037,0.000009207161,0.0003532872,0.0000995176,0.0003942882,0.00012994,0.0148436,0.983583,0.00001809606],"study_design_candidate":"not_applicable","study_design_consensus":"not_applicable","genre_codex":"editorial","genre_gemma":"editorial","genre_scores_codex":[0.001513104,0.1114024,0.1207027,0.1087616,0.5509495,0.0004921966,0.001412835,0.001771003,0.1029947],"genre_scores_gemma":[0.01003301,0.120644,0.05775074,0.03364803,0.6326497,0.0007762572,0.003678014,0.002526621,0.1382937],"genre_candidate":"editorial","genre_consensus":"editorial","teacher_disagreement_score":0.03282798,"threshold_uncertainty_score":0.1098205,"prediction_status":"machine_predicted_unvalidated"},"labels":[],"label_agreement":null},{"id":"W3173162544","doi":"10.1145/3567592","title":"Neural Machine Translation for Low-resource Languages: A Survey","year":2022,"lang":"en","type":"review","venue":"ACM Computing Surveys","topic":"Natural Language Processing Techniques","field":"Computer Science","cited_by":278,"is_retracted":false,"has_abstract":true,"routes":{"ca_aff":true,"ca_fund":false,"ca_venue":false,"about_ca":false},"ca_institutions":"University of Toronto","funders":"","keywords":"Computer science; Machine translation; Unavailability; Resource (disambiguation); Artificial intelligence; Natural language processing","authors":[{"name":"Surangika Ranathunga","is_ca":false},{"name":"En-Shiun Annie Lee","is_ca":true},{"name":"Marjana Prifti Skënduli","is_ca":false},{"name":"Ravi Shekhar","is_ca":false},{"name":"Mehreen Alam","is_ca":false},{"name":"Rishemjit Kaur","is_ca":false}],"retraction":null,"screen_n_in":null,"score":{"opus":0.09391014352774846,"gpt":0.3746552444027698,"spread":0.2807451008750214,"validation_status":"score_only:v0-immature-baseline"},"prediction":{"model_version":"metacan-v3-hybrid-931329e0061c","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.001907713,0.0008300945,0.001049091,0.004263788,0.0004510614,0.00152649,0.001200605,0.001039802,0.006815796],"category_scores_gemma":[0.006060451,0.0005702084,0.0007054436,0.007124552,0.0005747785,0.003628692,0.0008576522,0.001177328,0.005881336],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0007478356,"about_ca_system_score_gemma":0.001731699,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.001825927,"about_ca_topic_score_gemma":0.002706669,"domain_scores_codex":[0.9990159,0.0002339892,0.0001107914,0.0002044162,0.0003804414,0.00005444073],"domain_scores_gemma":[0.9961181,0.002607197,0.0001690322,0.0002086711,0.0008352168,0.00006182407],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"design_other","study_design_gemma":"not_applicable","study_design_scores_codex":[0.00003583272,0.00005713567,0.0005923346,0.006415778,0.00005756007,0.00006204106,0.00008196132,0.001580893,0.001182389,0.007909557,0.02381581,0.9582087],"study_design_scores_gemma":[0.00002291455,0.0002412144,0.002992401,0.003974381,0.0001669343,0.001284357,0.0003196609,0.01457196,0.00375283,0.01966535,0.9529429,0.00006499683],"study_design_candidate":"not_applicable","study_design_consensus":null,"genre_codex":"review","genre_gemma":"review","genre_scores_codex":[0.002742281,0.9512467,0.02775577,0.002200238,0.00055868,0.00006708728,0.000326619,0.0004610536,0.01464153],"genre_scores_gemma":[0.01670905,0.9529551,0.02273917,0.0006850493,0.0009224348,0.00008187735,0.001035443,0.0001245692,0.004747341],"genre_candidate":"review","genre_consensus":"review","teacher_disagreement_score":0.006815796,"threshold_uncertainty_score":0.0228011,"prediction_status":"machine_predicted_unvalidated"},"labels":[],"label_agreement":null},{"id":"W4400531953","doi":"10.1145/3626772.3657878","title":"C-Pack: Packed Resources For General Chinese Embeddings","year":2024,"lang":"en","type":"article","venue":"","topic":"Natural Language Processing Techniques","field":"Computer Science","cited_by":269,"is_retracted":false,"has_abstract":true,"routes":{"ca_aff":true,"ca_fund":false,"ca_venue":false,"about_ca":false},"ca_institutions":"Université de Montréal","funders":"","keywords":"Packed bed; Computer science; Chemistry; Chromatography","authors":[{"name":"Shitao Xiao","is_ca":false},{"name":"Zheng Liu","is_ca":false},{"name":"Peitian Zhang","is_ca":false},{"name":"Niklas Muennighoff","is_ca":false},{"name":"Defu Lian","is_ca":false},{"name":"Jian‐Yun Nie","is_ca":true}],"retraction":null,"screen_n_in":null,"score":{"opus":0.007724179664062534,"gpt":0.297602640136452,"spread":0.2898784604723895,"validation_status":"score_only:v0-immature-baseline"},"prediction":{"model_version":"metacan-v3-hybrid-931329e0061c","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.002541451,0.004819568,0.001697718,0.004080434,0.001157337,0.002799419,0.005478612,0.001612342,0.07963978],"category_scores_gemma":[0.02182814,0.002098715,0.002135489,0.006053665,0.0009586288,0.008683732,0.00747789,0.00336383,0.08575033],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.001441466,"about_ca_system_score_gemma":0.00406317,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.01034919,"about_ca_topic_score_gemma":0.0180387,"domain_scores_codex":[0.9976087,0.0004372998,0.0003025433,0.0005718104,0.0007512415,0.0003284097],"domain_scores_gemma":[0.9941643,0.001089242,0.0001858715,0.00295329,0.00127923,0.0003281335],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"not_applicable","study_design_gemma":"not_applicable","study_design_scores_codex":[0.0004009086,0.0002062394,0.001878065,0.001149854,0.0001057096,0.0002272533,0.0001893715,0.009778254,0.003550026,0.01259854,0.795925,0.1739907],"study_design_scores_gemma":[0.0005327882,0.0003589093,0.003054652,0.0004608603,0.0001136297,0.0006574704,0.000258019,0.1706529,0.03296746,0.06097267,0.7296016,0.0003689622],"study_design_candidate":"not_applicable","study_design_consensus":"not_applicable","genre_codex":"software","genre_gemma":"dataset","genre_scores_codex":[0.01357489,0.002095682,0.2957858,0.001150839,0.001789922,0.001046211,0.2654861,0.3832396,0.03583102],"genre_scores_gemma":[0.04240028,0.001100844,0.2562245,0.0009530527,0.0002409523,0.00262001,0.6304893,0.04987352,0.01609752],"genre_candidate":"dataset","genre_consensus":null,"teacher_disagreement_score":0.07963978,"threshold_uncertainty_score":0.2664217,"prediction_status":"machine_predicted_unvalidated"},"labels":[],"label_agreement":null},{"id":"W2143564602","doi":"10.3115/1220175.1220271","title":"An end-to-end discriminative approach to machine translation","year":2006,"lang":"en","type":"article","venue":"","topic":"Natural Language Processing Techniques","field":"Computer Science","cited_by":265,"is_retracted":false,"has_abstract":true,"routes":{"ca_aff":false,"ca_fund":true,"ca_venue":false,"about_ca":false},"ca_institutions":"","funders":"Fonds Québécois de la Recherche sur la Nature et les Technologies; Microsoft Research","keywords":"Discriminative model; Computer science; Artificial intelligence; Machine translation; Phrase; Heuristic; Feature (linguistics); Machine learning; Translation (biology); Process (computing); Decoding methods; Perceptron; Natural language processing; Pattern recognition (psychology); Artificial neural network; Algorithm","authors":[{"name":"Percy Liang","is_ca":false},{"name":"Alexandre Bouchard‐Côté","is_ca":false},{"name":"Dan Klein","is_ca":false},{"name":"Ben Taskar","is_ca":false}],"retraction":null,"screen_n_in":null,"score":{"opus":0.01802462325358629,"gpt":0.2828060711697341,"spread":0.2647814479161478,"validation_status":"score_only:v0-immature-baseline"},"prediction":{"model_version":"metacan-v3-hybrid-931329e0061c","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.001806251,0.001010403,0.001133973,0.001056044,0.001001364,0.00137644,0.002034458,0.00133933,0.005596932],"category_scores_gemma":[0.00587615,0.0006056253,0.0005107394,0.002023786,0.0007910636,0.002901812,0.002084572,0.002260582,0.005932039],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0004983838,"about_ca_system_score_gemma":0.001042049,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.001236069,"about_ca_topic_score_gemma":0.00381234,"domain_scores_codex":[0.9981565,0.0006205077,0.0001186994,0.000476714,0.0005096754,0.0001179148],"domain_scores_gemma":[0.9969982,0.0007181411,0.0002068761,0.001217866,0.000743375,0.0001154519],"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":[0.0002947923,0.0005040477,0.001735316,0.0002662379,0.00008628405,0.0002677279,0.0002227099,0.04578188,0.0459516,0.02562116,0.02086279,0.8584054],"study_design_scores_gemma":[0.00004806734,0.0002830554,0.0009719417,0.00002534126,0.00004562788,0.0008818382,0.0000601411,0.8865778,0.03790992,0.04795974,0.02516617,0.00007026581],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.004581206,0.0001061226,0.9872096,0.0001899223,0.00005145407,0.00004598395,0.0001732613,0.005091707,0.00255082],"genre_scores_gemma":[0.2071404,0.0002334057,0.7765785,0.0005928919,0.0001819965,0.0001559364,0.001997849,0.0008571449,0.01226188],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.005596932,"threshold_uncertainty_score":0.01872361,"prediction_status":"machine_predicted_unvalidated"},"labels":[],"label_agreement":null},{"id":"W2531882892","doi":"10.1111/cogs.12414","title":"Grammaticality, Acceptability, and Probability: A Probabilistic View of Linguistic Knowledge","year":2016,"lang":"en","type":"article","venue":"Cognitive Science","topic":"Natural Language Processing Techniques","field":"Computer Science","cited_by":261,"is_retracted":false,"has_abstract":true,"routes":{"ca_aff":false,"ca_fund":true,"ca_venue":false,"about_ca":false},"ca_institutions":"","funders":"Menzies Centre for Australian Studies, King's College London, University of London; Economic and Social Research Council; University of Toronto; Chalmers Tekniska Högskola; University of Reading; Universiteit van Amsterdam; Weizmann Institute of Science; Göteborgs Universitet; University of Edinburgh; King's College London","keywords":"Grammaticality; Probabilistic logic; Natural language processing; Artificial intelligence; Computer science; Sentence; Natural language; Set (abstract data type); Linguistics; Grammar","authors":[{"name":"Jey Han Lau","is_ca":false},{"name":"Shalom Lappin","is_ca":false}],"retraction":null,"screen_n_in":null,"score":{"opus":0.03081905117699348,"gpt":0.329199145905938,"spread":0.2983800947289446,"validation_status":"score_only:v0-immature-baseline"},"prediction":{"model_version":"metacan-v3-hybrid-931329e0061c","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.005987735,0.000674979,0.00112203,0.002697812,0.001400684,0.006310383,0.002173721,0.002527044,0.002606207],"category_scores_gemma":[0.04243263,0.0009142011,0.001404399,0.001637202,0.01418908,0.01670275,0.003812434,0.005449856,0.0003916425],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.001783516,"about_ca_system_score_gemma":0.001076363,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.004221367,"about_ca_topic_score_gemma":0.002389687,"domain_scores_codex":[0.9942795,0.002520929,0.0002330107,0.001124084,0.001599709,0.0002427426],"domain_scores_gemma":[0.9661623,0.02468512,0.002900643,0.003241659,0.002064391,0.0009457276],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"theoretical_or_conceptual","study_design_gemma":"theoretical_or_conceptual","study_design_scores_codex":[0.0002663161,0.0001295822,0.01839811,0.0003064181,0.0001983538,0.0002532915,0.006592702,0.0286567,0.005463217,0.8735008,0.001884814,0.06434987],"study_design_scores_gemma":[0.00001428229,0.00005454635,0.007579572,0.00004506372,0.00003366784,0.0002121968,0.0002892359,0.03810593,0.0006859952,0.950958,0.001955048,0.00006639712],"study_design_candidate":"theoretical_or_conceptual","study_design_consensus":"theoretical_or_conceptual","genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.1685991,0.001610315,0.7824072,0.015746,0.0001447185,0.00007291646,0.0003769823,0.000561556,0.03048122],"genre_scores_gemma":[0.9425486,0.0004838172,0.05424574,0.0008245838,0.0002485355,0.0001022444,0.000192483,0.0001386959,0.001215269],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.006310383,"threshold_uncertainty_score":0.03166652,"prediction_status":"machine_predicted_unvalidated"},"labels":[],"label_agreement":null},{"id":"W2092527610","doi":"10.1162/089120102760173625","title":"Near-Synonymy and Lexical Choice","year":2002,"lang":"en","type":"article","venue":"Computational Linguistics","topic":"Natural Language Processing Techniques","field":"Computer Science","cited_by":258,"is_retracted":false,"has_abstract":true,"routes":{"ca_aff":true,"ca_fund":false,"ca_venue":false,"about_ca":false},"ca_institutions":"University of Toronto","funders":"","keywords":"Computer science; Natural language processing; Denotation (semiotics); Artificial intelligence; Context (archaeology); Lexicon; Lexical choice; Set (abstract data type); Selection (genetic algorithm); Lexical semantics; Ontology; Lexical item; Linguistics","authors":[{"name":"Philip Edmonds","is_ca":false},{"name":"Graeme Hirst","is_ca":true}],"retraction":null,"screen_n_in":null,"score":{"opus":0.02487434097093336,"gpt":0.2844013043204678,"spread":0.2595269633495345,"validation_status":"score_only:v0-immature-baseline"},"prediction":{"model_version":"metacan-v3-hybrid-931329e0061c","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.002068811,0.00067599,0.001127268,0.002508612,0.001653466,0.004144925,0.003154934,0.001785794,0.007152132],"category_scores_gemma":[0.007947455,0.0007584203,0.002976573,0.00280112,0.004996182,0.01306055,0.00381253,0.002213376,0.001530941],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.002397544,"about_ca_system_score_gemma":0.001668088,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.003954945,"about_ca_topic_score_gemma":0.003367992,"domain_scores_codex":[0.9967014,0.001005468,0.0003051002,0.001024087,0.0007608715,0.0002030073],"domain_scores_gemma":[0.9966118,0.001731492,0.0003599847,0.0007326696,0.0004006295,0.0001634018],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"theoretical_or_conceptual","study_design_gemma":"simulation_or_modeling","study_design_scores_codex":[0.00003662437,0.00002398611,0.000564302,0.00007986676,0.00003870005,0.000151933,0.0005940836,0.02199537,0.001162462,0.9567356,0.000641158,0.01797595],"study_design_scores_gemma":[0.00001469278,0.00001766167,0.0001630168,0.00002277126,0.00001717957,0.00009750773,0.0001279735,0.09206878,0.0006908448,0.9015422,0.005215084,0.00002235286],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.0215734,0.0002085187,0.9677023,0.0007176159,0.0000503013,0.0001036421,0.0002081746,0.000194744,0.009241168],"genre_scores_gemma":[0.4640301,0.0002978935,0.5263891,0.0002972171,0.00007738081,0.0004240198,0.0006255143,0.0001373309,0.007721356],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.007152132,"threshold_uncertainty_score":0.02392632,"prediction_status":"machine_predicted_unvalidated"},"labels":[],"label_agreement":null},{"id":"W2164948578","doi":"10.1162/089120103321337458","title":"Word Reordering and a Dynamic Programming Beam Search Algorithm for Statistical Machine Translation","year":2003,"lang":"en","type":"article","venue":"Computational Linguistics","topic":"Natural Language Processing Techniques","field":"Computer Science","cited_by":257,"is_retracted":false,"has_abstract":true,"routes":{"ca_aff":false,"ca_fund":false,"ca_venue":false,"about_ca":true},"ca_institutions":"","funders":"","keywords":"Computer science; Word (group theory); Task (project management); Machine translation; Vocabulary; Beam search; Word error rate; Sentence; Translation (biology); Artificial intelligence; Natural language processing; Dynamic programming; Search algorithm; Speech recognition; Set (abstract data type); Algorithm; Linguistics; Programming language","authors":[{"name":"Christoph Tillmann","is_ca":false},{"name":"Hermann Ney","is_ca":false}],"retraction":null,"screen_n_in":null,"score":{"opus":0.01830414375166094,"gpt":0.319656416215049,"spread":0.3013522724633881,"validation_status":"score_only:v0-immature-baseline"},"prediction":{"model_version":"metacan-v3-hybrid-931329e0061c","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.002154728,0.001103758,0.001403458,0.001525322,0.0007651085,0.001068281,0.001575622,0.001337416,0.00535028],"category_scores_gemma":[0.006271593,0.0009679737,0.001090812,0.003113655,0.001133488,0.002148295,0.001249659,0.001946921,0.002357747],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0007492857,"about_ca_system_score_gemma":0.001717256,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.003122549,"about_ca_topic_score_gemma":0.004319443,"domain_scores_codex":[0.9980897,0.0009716555,0.0001142469,0.000310175,0.0004245213,0.0000896112],"domain_scores_gemma":[0.9977291,0.001539801,0.0001121295,0.0002237218,0.0003580615,0.00003724744],"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":[0.0002129682,0.0001214303,0.0004349302,0.0002825667,0.000163471,0.0001514684,0.0002946887,0.2715408,0.01063096,0.1135885,0.008608317,0.5939699],"study_design_scores_gemma":[0.0000636851,0.00006043009,0.0001053877,0.00001891571,0.00002505707,0.00007336041,0.00002686806,0.9234301,0.002892099,0.06645893,0.006819246,0.00002590173],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"methods","genre_scores_codex":[0.001408353,0.0001378291,0.9973579,0.00006170898,0.00002311057,0.0000342612,0.0000313814,0.0003778893,0.0005675186],"genre_scores_gemma":[0.02167759,0.0001711928,0.9760864,0.0001059655,0.00003259958,0.0002771794,0.0002047074,0.0002275124,0.001216878],"genre_candidate":"methods","genre_consensus":"methods","teacher_disagreement_score":0.00535028,"threshold_uncertainty_score":0.0178985,"prediction_status":"machine_predicted_unvalidated"},"labels":[],"label_agreement":null},{"id":"W2132001515","doi":"10.3115/1626355.1626372","title":"Mixture-model adaptation for SMT","year":2007,"lang":"en","type":"article","venue":"","topic":"Natural Language Processing Techniques","field":"Computer Science","cited_by":253,"is_retracted":false,"has_abstract":true,"routes":{"ca_aff":true,"ca_fund":false,"ca_venue":false,"about_ca":false},"ca_institutions":"National Research Council Canada","funders":"","keywords":"Computer science; Machine translation; Granularity; Adaptation (eye); BLEU; Translation (biology); Domain adaptation; Log-linear model; Artificial intelligence; Baseline (sea); Language model; Natural language processing; Algorithm; Linear model; Machine learning; Programming language","authors":[{"name":"George Foster","is_ca":true},{"name":"Roland Kühn","is_ca":true}],"retraction":null,"screen_n_in":null,"score":{"opus":0.02364179146064855,"gpt":0.3049751683204618,"spread":0.2813333768598132,"validation_status":"score_only:v0-immature-baseline"},"prediction":{"model_version":"metacan-v3-hybrid-931329e0061c","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.002298851,0.001415706,0.001237829,0.00116692,0.0006590079,0.00120203,0.001778068,0.001424476,0.007811815],"category_scores_gemma":[0.008119074,0.0008529921,0.001791203,0.002086868,0.0005953886,0.002339152,0.001960437,0.003313594,0.008248528],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0008795568,"about_ca_system_score_gemma":0.0009200506,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.003681744,"about_ca_topic_score_gemma":0.005250933,"domain_scores_codex":[0.9981043,0.0008678616,0.0001105829,0.0004091107,0.000421744,0.00008634387],"domain_scores_gemma":[0.9977738,0.001120694,0.00009361433,0.0005488326,0.0004030459,0.00005991528],"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":[0.0003308917,0.0001629483,0.0008599714,0.0003346719,0.0005486141,0.0001741034,0.0001817471,0.3547237,0.02255477,0.02843252,0.01777523,0.5739208],"study_design_scores_gemma":[0.00001636182,0.00003846316,0.0002475987,0.00001573685,0.0000439665,0.0001146629,0.00001403522,0.963197,0.006985798,0.01891839,0.01037139,0.00003661113],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.001307785,0.000264961,0.9946031,0.00009876583,0.00008694438,0.00003837477,0.0001385937,0.002433228,0.00102822],"genre_scores_gemma":[0.1298004,0.0008173825,0.8548298,0.0003309304,0.0002440058,0.0005011178,0.002056871,0.002110157,0.009309384],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.007811815,"threshold_uncertainty_score":0.02613312,"prediction_status":"machine_predicted_unvalidated"},"labels":[],"label_agreement":null},{"id":"W4285084440","doi":"10.1038/s41587-022-01369-0","title":"Standardized annotation of translated open reading frames","year":2022,"lang":"en","type":"letter","venue":"Nature Biotechnology","topic":"Natural Language Processing Techniques","field":"Computer Science","cited_by":250,"is_retracted":false,"has_abstract":true,"routes":{"ca_aff":true,"ca_fund":true,"ca_venue":false,"about_ca":false},"ca_institutions":"Université de Sherbrooke","funders":"National Cancer Institute; National Human Genome Research Institute; National Institute on Aging; RIKEN; Agència de Gestió d'Ajuts Universitaris i de Recerca; Medical Research Council; Canadian Institutes of Health Research; Directorate for Biological Sciences; National Institutes of Health; European Molecular Biology Laboratory; Russian Science Foundation; Institució Catalana de Recerca i Estudis Avançats; National Institute of General Medical Sciences; University College Cork; University of Leeds; Universitat Pompeu Fabra; Fondation Leducq; Stowers Institute for Medical Research; Australian Government; Searle Scholars Program; National Science Foundation; Science Foundation Ireland; Biotechnology and Biological Sciences Research Council; Université de Sherbrooke; Wellcome Trust; European Commission; University of California, Irvine; Broad Institute; University of Pittsburgh; Howard Hughes Medical Institute; Staatssekretariat für Bildung, Forschung und Innovation; Musella Foundation For Brain Tumor Research and Information; Agencia Estatal de Investigación; Computer Science and Artificial Intelligence Laboratory, Massachusetts Institute of Technology; Alex's Lemonade Stand Foundation for Childhood Cancer; Yale University","keywords":"Annotation; Open reading frame; Reading (process); Computer science; Information retrieval; Computational biology; Natural language processing; Biology; Artificial intelligence; Linguistics; Genetics; Philosophy; Peptide sequence","authors":[{"name":"Jonathan M. Mudge","is_ca":false},{"name":"Jorge Ruiz‐Orera","is_ca":false},{"name":"John R. Prensner","is_ca":false},{"name":"Marie A. Brunet","is_ca":true},{"name":"Ferriol Calvet","is_ca":false},{"name":"Irwin Jungreis","is_ca":false},{"name":"José M. González","is_ca":false},{"name":"Michele Magrane","is_ca":false},{"name":"Thomas F. Martínez","is_ca":false},{"name":"Jana Felicitas Schulz","is_ca":false},{"name":"Yucheng Yang","is_ca":false},{"name":"M. Mar Albà","is_ca":false},{"name":"Julie L. Aspden","is_ca":false},{"name":"Pavel V. Baranov","is_ca":false},{"name":"Ariel Bazzini","is_ca":false},{"name":"Elspeth A. Bruford","is_ca":false},{"name":"María Martin","is_ca":false},{"name":"Lorenzo Calviello","is_ca":false},{"name":"Anne‐Ruxandra Carvunis","is_ca":false},{"name":"Jin Chen","is_ca":false},{"name":"Juan Pablo Couso","is_ca":false},{"name":"Eric W. Deutsch","is_ca":false},{"name":"Paul Flicek","is_ca":false},{"name":"Adam Frankish","is_ca":false},{"name":"Mark Gerstein","is_ca":false},{"name":"Norbert Hübner","is_ca":false},{"name":"Nicholas T. Ingolia","is_ca":false},{"name":"Manolis Kellis","is_ca":false},{"name":"Gerben Menschaert","is_ca":false},{"name":"Robert L. Moritz","is_ca":false},{"name":"Uwe Ohler","is_ca":false},{"name":"Xavier Roucou","is_ca":true},{"name":"Alan Saghatelian","is_ca":false},{"name":"Jonathan S. Weissman","is_ca":false},{"name":"Sebastiaan van Heesch","is_ca":false}],"retraction":null,"screen_n_in":null,"score":{"opus":0.009975389999421257,"gpt":0.2940358937373858,"spread":0.2840605037379645,"validation_status":"score_only:v0-immature-baseline"},"prediction":{"model_version":"metacan-v3-hybrid-931329e0061c","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.0071533,0.0007279761,0.0008087646,0.002858799,0.001162072,0.002156815,0.00156726,0.002574262,0.003296503],"category_scores_gemma":[0.01413657,0.0004299335,0.0005875914,0.002580398,0.001446949,0.00165101,0.001346365,0.005032898,0.009486219],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.002477055,"about_ca_system_score_gemma":0.00318193,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.00145654,"about_ca_topic_score_gemma":0.003699841,"domain_scores_codex":[0.9929289,0.002218407,0.001180352,0.000928584,0.002367863,0.0003758318],"domain_scores_gemma":[0.9837516,0.004778243,0.001850779,0.003373828,0.005642054,0.0006035283],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"not_applicable","study_design_gemma":"not_applicable","study_design_scores_codex":[0.0003485095,0.0001093698,0.003219853,0.001190983,0.00003970233,0.001974623,0.0005121077,0.0007346225,0.08339601,0.06544748,0.4837328,0.3592939],"study_design_scores_gemma":[0.00002464286,0.0001050502,0.002354124,0.0003061919,0.00002028027,0.001242511,0.0001709068,0.001626858,0.02074423,0.01449096,0.9588595,0.00005480846],"study_design_candidate":"not_applicable","study_design_consensus":"not_applicable","genre_codex":"methods","genre_gemma":"methods","genre_scores_codex":[0.0322214,0.05416159,0.3994699,0.3290198,0.0579285,0.001455183,0.02582556,0.007636494,0.09228159],"genre_scores_gemma":[0.08831321,0.03407707,0.5939028,0.140617,0.02244526,0.001804017,0.06970236,0.00266763,0.04647072],"genre_candidate":"methods","genre_consensus":"methods","teacher_disagreement_score":0.0071533,"threshold_uncertainty_score":0.03783071,"prediction_status":"machine_predicted_unvalidated"},"labels":[],"label_agreement":null},{"id":"W2200913422","doi":"10.33011/lilt.v9i.1321","title":"Frege in Space: A Program for Compositional Distributional Semantics","year":2014,"lang":"en","type":"article","venue":"Linguistic Issues in Language Technology","topic":"Natural Language Processing Techniques","field":"Computer Science","cited_by":245,"is_retracted":false,"has_abstract":true,"routes":{"ca_aff":false,"ca_fund":true,"ca_venue":false,"about_ca":false},"ca_institutions":"","funders":"Atomic Energy of Canada Limited","keywords":"Principle of compositionality; Meaning (existential); Computer science; Semantics (computer science); Distributional semantics; Lexicon; Lexicology; Linguistics; Syntax; Formal semantics (linguistics); Computational semantics; Lexical semantics; Natural language processing; Cognitive semantics; Artificial intelligence; Space (punctuation); Function (biology); Operational semantics; Lexical item; Programming language; Cognition; Epistemology; Psychology; Philosophy","authors":[{"name":"Marco Baroni","is_ca":false},{"name":"Raffaella Bernardi","is_ca":false},{"name":"Roberto Zamparelli","is_ca":false}],"retraction":null,"screen_n_in":null,"score":{"opus":0.005157375645620031,"gpt":0.3124737237818904,"spread":0.3073163481362703,"validation_status":"score_only:v0-immature-baseline"},"prediction":{"model_version":"metacan-v3-hybrid-931329e0061c","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.002777147,0.001122144,0.0008740491,0.002436698,0.001777807,0.003280514,0.002176889,0.00147287,0.02430371],"category_scores_gemma":[0.008366975,0.0007652715,0.002879212,0.001827618,0.003167751,0.007947937,0.00470953,0.002429848,0.004473676],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.001451078,"about_ca_system_score_gemma":0.001432921,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.002599016,"about_ca_topic_score_gemma":0.004262549,"domain_scores_codex":[0.9988301,0.0004622464,0.00007101303,0.0002782278,0.0002633,0.0000951745],"domain_scores_gemma":[0.998419,0.001179206,0.00004992482,0.0001832388,0.0001252592,0.00004345725],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"theoretical_or_conceptual","study_design_gemma":"theoretical_or_conceptual","study_design_scores_codex":[0.00008765875,0.00003896751,0.0003612892,0.0001535743,0.0000385542,0.0001427172,0.0008111135,0.01158823,0.0008100831,0.9061144,0.00917208,0.07068125],"study_design_scores_gemma":[0.00003324704,0.00001700348,0.00006825338,0.00004579401,0.00001723505,0.00008604847,0.0001550164,0.07971904,0.001090519,0.8770882,0.04165698,0.00002251457],"study_design_candidate":"theoretical_or_conceptual","study_design_consensus":"theoretical_or_conceptual","genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.001364994,0.00006410951,0.9888654,0.0003235408,0.00005078023,0.00005170939,0.0002131842,0.004074375,0.00499194],"genre_scores_gemma":[0.06729056,0.0002360305,0.9206218,0.0003240953,0.0001312704,0.0004584554,0.0006418788,0.003126602,0.00716931],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.02430371,"threshold_uncertainty_score":0.08130401,"prediction_status":"machine_predicted_unvalidated"},"labels":[],"label_agreement":null},{"id":"W2026222783","doi":"10.7202/1025047ar","title":"Translation Skill-Sets in a Machine-Translation Age","year":2014,"lang":"en","type":"article","venue":"Meta Journal des traducteurs","topic":"Natural Language Processing Techniques","field":"Computer Science","cited_by":241,"is_retracted":false,"has_abstract":true,"routes":{"ca_aff":false,"ca_fund":false,"ca_venue":true,"about_ca":false},"ca_institutions":"","funders":"","keywords":"Terminology; Computer science; Machine translation; Space (punctuation); Translation (biology); Computer-assisted translation; Function (biology); Natural language processing; Artificial intelligence; Linguistics","authors":[{"name":"Anthony Pym","is_ca":false}],"retraction":null,"screen_n_in":null,"score":{"opus":0.03194641009699978,"gpt":0.290480704647114,"spread":0.2585342945501142,"validation_status":"score_only:v0-immature-baseline"},"prediction":{"model_version":"metacan-v3-hybrid-931329e0061c","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.01409449,0.0008576661,0.0008013857,0.003613642,0.003613626,0.00884664,0.001984353,0.003577325,0.009901158],"category_scores_gemma":[0.02777594,0.0005811451,0.0005274857,0.001353585,0.01994571,0.01530776,0.008299931,0.005799409,0.00449317],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.003736125,"about_ca_system_score_gemma":0.004580376,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.001777211,"about_ca_topic_score_gemma":0.001164995,"domain_scores_codex":[0.9919612,0.003602993,0.0006158871,0.00167442,0.001710941,0.0004345664],"domain_scores_gemma":[0.9750589,0.01622995,0.001038645,0.002908878,0.003346544,0.001417022],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"theoretical_or_conceptual","study_design_gemma":"theoretical_or_conceptual","study_design_scores_codex":[0.0001540955,0.000269348,0.00551692,0.0007927385,0.00004679459,0.0004960776,0.06719696,0.00181592,0.003861609,0.5998092,0.02039166,0.2996487],"study_design_scores_gemma":[0.00003549929,0.0001815564,0.006855746,0.0007318599,0.00001696126,0.000739403,0.01544879,0.004000565,0.002215109,0.7712152,0.1984632,0.00009619231],"study_design_candidate":"theoretical_or_conceptual","study_design_consensus":"theoretical_or_conceptual","genre_codex":"other","genre_gemma":"empirical","genre_scores_codex":[0.1780721,0.02613546,0.2920385,0.1463575,0.002354448,0.0004472734,0.0005527867,0.00142121,0.3526208],"genre_scores_gemma":[0.7655885,0.007965048,0.1659494,0.01388411,0.001693933,0.0007077521,0.0005810594,0.0004925122,0.04313753],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.01409449,"threshold_uncertainty_score":0.07453966,"prediction_status":"machine_predicted_unvalidated"},"labels":[],"label_agreement":null},{"id":"W2152808281","doi":"10.1109/tnn.2007.912312","title":"Adaptive Importance Sampling to Accelerate Training of a Neural Probabilistic Language Model","year":2008,"lang":"en","type":"article","venue":"IEEE Transactions on Neural Networks","topic":"Natural Language Processing Techniques","field":"Computer Science","cited_by":233,"is_retracted":false,"has_abstract":true,"routes":{"ca_aff":true,"ca_fund":false,"ca_venue":false,"about_ca":false},"ca_institutions":"Université de Montréal","funders":"","keywords":"Computer science; Language model; Artificial neural network; Vocabulary; Probabilistic neural network; Speedup; Artificial intelligence; Probabilistic logic; Computation; Machine learning; Feedforward neural network; Sampling (signal processing); Statistical model; Training (meteorology); Backpropagation; Importance sampling; Time delay neural network; Algorithm; Statistics; Mathematics; Monte Carlo method","authors":[{"name":"Yoshua Bengio","is_ca":true},{"name":"Jean-Sébastien Senécal","is_ca":true}],"retraction":null,"screen_n_in":null,"score":{"opus":0.0677840225644307,"gpt":0.2919126299227441,"spread":0.2241286073583134,"validation_status":"score_only:v0-immature-baseline"},"prediction":{"model_version":"metacan-v3-hybrid-931329e0061c","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.001585378,0.0005938052,0.0007230479,0.0004931991,0.0002870628,0.000416031,0.001440164,0.0007244152,0.001907883],"category_scores_gemma":[0.006871984,0.0005024433,0.0004509148,0.0006185902,0.0004797015,0.001571464,0.00104709,0.001811918,0.0005297738],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0005440061,"about_ca_system_score_gemma":0.0009135033,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.004947551,"about_ca_topic_score_gemma":0.007286832,"domain_scores_codex":[0.9993852,0.0002446672,0.00003064425,0.00009257408,0.0001869738,0.00005993163],"domain_scores_gemma":[0.9979785,0.001421674,0.00007275648,0.0002142424,0.0002505785,0.00006222086],"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":[0.0002780461,0.0001704341,0.001164615,0.00009962964,0.00005380576,0.0001457789,0.0001291847,0.6906656,0.01374676,0.02257713,0.002846483,0.2681226],"study_design_scores_gemma":[0.000003942233,0.000007090905,0.00003231261,8.368856e-7,0.000001557699,0.00000713231,0.000001417201,0.997331,0.0007044728,0.001771098,0.0001379753,0.000001223942],"study_design_candidate":"simulation_or_modeling","study_design_consensus":"simulation_or_modeling","genre_codex":"methods","genre_gemma":"methods","genre_scores_codex":[0.01263745,0.0001067266,0.9860241,0.00009055774,0.00003229607,0.00002218457,0.00001599635,0.0006980363,0.0003725924],"genre_scores_gemma":[0.3584071,0.0001908351,0.6390409,0.0001295552,0.0001077151,0.0001352105,0.0002067091,0.0001903799,0.0015916],"genre_candidate":"methods","genre_consensus":"methods","teacher_disagreement_score":0.004947551,"threshold_uncertainty_score":0.009837508,"prediction_status":"machine_predicted_unvalidated"},"labels":[],"label_agreement":null},{"id":"W59585178","doi":"","title":"Proceedings of the conference on Human Language Technology and Empirical Methods in Natural Language Processing","year":2005,"lang":"en","type":"article","venue":"","topic":"Natural Language Processing Techniques","field":"Computer Science","cited_by":231,"is_retracted":false,"has_abstract":true,"routes":{"ca_aff":false,"ca_fund":false,"ca_venue":false,"about_ca":true},"ca_institutions":"","funders":"","keywords":"Language technology; Computer science; Human language; Computational linguistics; Linguistics; Natural language; Artificial intelligence; Natural language processing; Comprehension approach","authors":[{"name":"Raymond J. Mooney","is_ca":false}],"retraction":null,"screen_n_in":null,"score":{"opus":0.03212528760022702,"gpt":0.4063860398749459,"spread":0.3742607522747188,"validation_status":"score_only:v0-immature-baseline"},"prediction":{"model_version":"metacan-v3-hybrid-931329e0061c","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.0270581,0.001017946,0.001620435,0.002800975,0.001642589,0.009266915,0.001848959,0.002739247,0.05195588],"category_scores_gemma":[0.04601075,0.0007413501,0.001177539,0.00244539,0.005367933,0.007511328,0.004081606,0.006003789,0.0120775],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.003291567,"about_ca_system_score_gemma":0.006286353,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.004591037,"about_ca_topic_score_gemma":0.005836438,"domain_scores_codex":[0.9848462,0.009132113,0.0008828907,0.001751459,0.003039558,0.0003477598],"domain_scores_gemma":[0.9612511,0.02490112,0.0007457624,0.004768686,0.006037326,0.002296054],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"not_applicable","study_design_gemma":"not_applicable","study_design_scores_codex":[0.0002447568,0.0001479839,0.001745458,0.0007170631,0.0001883811,0.0001838887,0.001184193,0.0006651462,0.001336662,0.09778519,0.5751151,0.3206863],"study_design_scores_gemma":[0.00004144918,0.00004529316,0.001634422,0.0004478576,0.00004482815,0.0002383832,0.0003708107,0.001635021,0.0006063615,0.05243725,0.9424552,0.00004300962],"study_design_candidate":"not_applicable","study_design_consensus":"not_applicable","genre_codex":"methods","genre_gemma":"other","genre_scores_codex":[0.008638581,0.1632279,0.4012544,0.1488839,0.07094289,0.001202936,0.00469626,0.002445932,0.1987073],"genre_scores_gemma":[0.08745792,0.09682527,0.3814164,0.02316629,0.04185366,0.002476557,0.01068284,0.002992604,0.3531284],"genre_candidate":"other","genre_consensus":null,"teacher_disagreement_score":0.05195588,"threshold_uncertainty_score":0.1738097,"prediction_status":"machine_predicted_unvalidated"},"labels":[],"label_agreement":null},{"id":"W3194817747","doi":"10.82308/39461","title":"Verb-stranding VP ellipsis : a cross-linguistic study","year":2005,"lang":"en","type":"article","venue":"eScholarship@McGill (McGill)","topic":"Natural Language Processing Techniques","field":"Computer Science","cited_by":226,"is_retracted":false,"has_abstract":true,"routes":{"ca_aff":false,"ca_fund":true,"ca_venue":false,"about_ca":false},"ca_institutions":"","funders":"McGill University","keywords":"Linguistics; Ellipsis (linguistics); Verb; Verb phrase ellipsis; Computer science; Artificial intelligence; Natural language processing; Modal verb; Philosophy","authors":[{"name":"Lotus Madelyn Goldberg","is_ca":false}],"retraction":null,"screen_n_in":null,"score":{"opus":0.02004930865468135,"gpt":0.2869970620228592,"spread":0.2669477533681779,"validation_status":"score_only:v0-immature-baseline"},"prediction":{"model_version":"metacan-v3-hybrid-931329e0061c","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.002534123,0.0004273853,0.0003267527,0.002011664,0.002446163,0.003115189,0.0006821365,0.001018975,0.005272744],"category_scores_gemma":[0.005833379,0.0003564993,0.0002755533,0.003432633,0.002503311,0.003989412,0.002746244,0.001161377,0.0006221549],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0009025317,"about_ca_system_score_gemma":0.0004515391,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.006592319,"about_ca_topic_score_gemma":0.009218864,"domain_scores_codex":[0.9986116,0.0006568338,0.0001012346,0.0002868223,0.0002174953,0.0001261194],"domain_scores_gemma":[0.9962134,0.002667077,0.0003584034,0.0003665739,0.0003141848,0.00008027395],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"qualitative","study_design_gemma":"observational","study_design_scores_codex":[0.0001368478,0.0004536629,0.07264341,0.0005778403,0.00009535474,0.007500843,0.7267287,0.0002644853,0.01468865,0.09231065,0.001966391,0.08263315],"study_design_scores_gemma":[0.00005109291,0.0004254205,0.3151063,0.0007045407,0.0002029147,0.01999392,0.4240917,0.002891431,0.009580446,0.02834792,0.1984122,0.0001921555],"study_design_candidate":"observational","study_design_consensus":null,"genre_codex":"empirical","genre_gemma":"empirical","genre_scores_codex":[0.9218153,0.002003461,0.006113565,0.0004954354,0.00007345645,0.00004734927,0.000119893,0.00003973406,0.06929176],"genre_scores_gemma":[0.9912117,0.0009851726,0.002078976,0.0001645885,0.00007228072,0.00003429163,0.0001334119,0.00005028204,0.005269275],"genre_candidate":"empirical","genre_consensus":"empirical","teacher_disagreement_score":0.006592319,"threshold_uncertainty_score":0.0176391,"prediction_status":"machine_predicted_unvalidated"},"labels":[],"label_agreement":null},{"id":"W2046436395","doi":"10.1162/089120100561755","title":"The Rhetorical Parsing of Unrestricted Texts: A Surface-based Approach","year":2000,"lang":"en","type":"article","venue":"Computational Linguistics","topic":"Natural Language Processing Techniques","field":"Computer Science","cited_by":221,"is_retracted":false,"has_abstract":true,"routes":{"ca_aff":false,"ca_fund":true,"ca_venue":false,"about_ca":false},"ca_institutions":"","funders":"Natural Sciences and Engineering Research Council of Canada; University of Toronto","keywords":"Rhetorical question; Computer science; Automatic summarization; Parsing; Natural language processing; Artificial intelligence; Linguistics; Context (archaeology); Natural language understanding; Natural language; Philosophy; History","authors":[{"name":"Daniel Marcu","is_ca":false}],"retraction":null,"screen_n_in":null,"score":{"opus":0.02046125002567074,"gpt":0.2820272355339836,"spread":0.2615659855083128,"validation_status":"score_only:v0-immature-baseline"},"prediction":{"model_version":"metacan-v3-hybrid-931329e0061c","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.00255761,0.0008841841,0.00107342,0.004258771,0.001415092,0.005042968,0.001637241,0.001328135,0.003956576],"category_scores_gemma":[0.01841367,0.0008874147,0.001124462,0.002547296,0.003182783,0.006781413,0.001728219,0.002274546,0.001929533],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0008654835,"about_ca_system_score_gemma":0.001369426,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.0006661028,"about_ca_topic_score_gemma":0.0009330409,"domain_scores_codex":[0.9977932,0.001156896,0.0001218225,0.0003718965,0.0004825299,0.00007371001],"domain_scores_gemma":[0.9881666,0.008296403,0.0006829425,0.001501563,0.001227065,0.0001252799],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"design_other","study_design_gemma":"bench_or_experimental","study_design_scores_codex":[0.0001796173,0.000179363,0.002702091,0.001293525,0.0001168052,0.0004081225,0.00441657,0.04570746,0.03728511,0.2927009,0.005571573,0.6094388],"study_design_scores_gemma":[0.00004978351,0.0001900954,0.002142844,0.0002429165,0.0001090402,0.0004639716,0.001334527,0.5034836,0.03129079,0.4276727,0.03292878,0.00009093446],"study_design_candidate":"bench_or_experimental","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.01376512,0.00039679,0.9803957,0.000404182,0.00003417741,0.0001192183,0.0002051501,0.001745015,0.002934744],"genre_scores_gemma":[0.1633496,0.0005380128,0.8322905,0.0001072196,0.00009788948,0.0001596084,0.001150515,0.0008248244,0.001481868],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.005042968,"threshold_uncertainty_score":0.01352608,"prediction_status":"machine_predicted_unvalidated"},"labels":[],"label_agreement":null},{"id":"W2548230849","doi":"10.1075/nlp.2.15mey","title":"Extracting knowledge-rich contexts for terminography","year":2001,"lang":"en","type":"book-chapter","venue":"Natural language processing","topic":"Natural Language Processing Techniques","field":"Computer Science","cited_by":215,"is_retracted":false,"has_abstract":true,"routes":{"ca_aff":true,"ca_fund":false,"ca_venue":false,"about_ca":false},"ca_institutions":"University of Ottawa","funders":"","keywords":"Construct (python library); Computer science; Paralanguage; Domain knowledge; Domain (mathematical analysis); Knowledge extraction; Field (mathematics); Context (archaeology); Knowledge management; Natural language processing; Data science; Artificial intelligence; Psychology; Communication; Geography; Mathematics","authors":[{"name":"Ingrid Meyer","is_ca":true}],"retraction":null,"screen_n_in":null,"score":{"opus":0.01563003246822913,"gpt":0.3028685290154244,"spread":0.2872384965471953,"validation_status":"score_only:v0-immature-baseline"},"prediction":{"model_version":"metacan-v3-hybrid-931329e0061c","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.001388135,0.0006908762,0.0005911522,0.00303333,0.001131591,0.002879164,0.00116295,0.0007922247,0.004759598],"category_scores_gemma":[0.005998624,0.0007256229,0.000913582,0.003312267,0.001411841,0.0072093,0.002355797,0.001868099,0.003611285],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.000786385,"about_ca_system_score_gemma":0.001068362,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.0005800364,"about_ca_topic_score_gemma":0.001760938,"domain_scores_codex":[0.9990559,0.0003599697,0.00009406748,0.000157749,0.0002960226,0.00003618616],"domain_scores_gemma":[0.9973213,0.001852538,0.00009813302,0.00042182,0.0002590098,0.00004717746],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"design_other","study_design_gemma":"not_applicable","study_design_scores_codex":[0.00006190898,0.00003753115,0.001043282,0.001316286,0.00004172568,0.0006070177,0.003521673,0.003542223,0.0102667,0.4065999,0.02007541,0.5528864],"study_design_scores_gemma":[0.00002108273,0.00003236941,0.001284014,0.001198116,0.00007026547,0.001368206,0.00097531,0.02688494,0.01558705,0.5650083,0.3875002,0.00007004268],"study_design_candidate":"not_applicable","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"methods","genre_scores_codex":[0.006694648,0.005613583,0.9614586,0.000645461,0.0002187884,0.0001626671,0.0004666967,0.001202827,0.02353679],"genre_scores_gemma":[0.03957,0.005195871,0.9443949,0.0001824101,0.0001314663,0.0002233798,0.001525029,0.0004938144,0.008283064],"genre_candidate":"methods","genre_consensus":"methods","teacher_disagreement_score":0.004759598,"threshold_uncertainty_score":0.01592249,"prediction_status":"machine_predicted_unvalidated"},"labels":[],"label_agreement":null},{"id":"W1848260265","doi":"","title":"Discriminative Instance Weighting for Domain Adaptation in Statistical Machine Translation","year":2010,"lang":"en","type":"article","venue":"NPARC","topic":"Natural Language Processing Techniques","field":"Computer Science","cited_by":211,"is_retracted":false,"has_abstract":true,"routes":{"ca_aff":false,"ca_fund":false,"ca_venue":true,"about_ca":false},"ca_institutions":"","funders":"","keywords":"Discriminative model; Weighting; Computer science; Domain adaptation; Machine translation; Artificial intelligence; Relevance (law); Granularity; Phrase; Natural language processing; Domain (mathematical analysis); Machine learning; Range (aeronautics); Adaptation (eye); Translation (biology); Pattern recognition (psychology); Mathematics; Engineering","authors":[{"name":"George Foster","is_ca":false},{"name":"Cyril Goutte","is_ca":false},{"name":"Roland Kühn","is_ca":false}],"retraction":null,"screen_n_in":null,"score":{"opus":0.01834908506549149,"gpt":0.2900663894007872,"spread":0.2717173043352957,"validation_status":"score_only:v0-immature-baseline"},"prediction":{"model_version":"metacan-v3-hybrid-931329e0061c","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.002754747,0.0009690283,0.001215262,0.001297385,0.0005009207,0.0009891014,0.001865944,0.001086342,0.003015335],"category_scores_gemma":[0.008819309,0.0005698251,0.00076605,0.00269182,0.0008518529,0.003116325,0.002029917,0.002578774,0.002322875],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0005895585,"about_ca_system_score_gemma":0.000628259,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.001602368,"about_ca_topic_score_gemma":0.002718308,"domain_scores_codex":[0.9982185,0.0009706583,0.00008846035,0.0003350178,0.0002885681,0.00009887249],"domain_scores_gemma":[0.9972121,0.001479107,0.0001664554,0.0008140229,0.0002417552,0.00008650998],"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":[0.0003409699,0.0002591989,0.001370993,0.0002381575,0.0001821059,0.0001385576,0.0001826834,0.1848678,0.02132192,0.03148202,0.008839818,0.7507758],"study_design_scores_gemma":[0.00002262174,0.00005781337,0.0003147529,0.00001263811,0.00002776221,0.0001027247,0.00002327797,0.9489197,0.006706495,0.03888143,0.00490518,0.00002573247],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.005499462,0.0004119468,0.9916489,0.00009135697,0.00005222323,0.00004304993,0.00008889403,0.0014313,0.0007329109],"genre_scores_gemma":[0.2685875,0.0008068564,0.7247429,0.0003175688,0.0003004571,0.0002843917,0.001105207,0.0007563533,0.003098784],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.003015335,"threshold_uncertainty_score":0.01456869,"prediction_status":"machine_predicted_unvalidated"},"labels":[],"label_agreement":null},{"id":"W179314280","doi":"","title":"Discriminative Reranking for Machine Translation","year":2004,"lang":"en","type":"article","venue":"","topic":"Natural Language Processing Techniques","field":"Computer Science","cited_by":207,"is_retracted":false,"has_abstract":true,"routes":{"ca_aff":true,"ca_fund":false,"ca_venue":false,"about_ca":false},"ca_institutions":"Simon Fraser University","funders":"","keywords":"Machine translation; Discriminative model; NIST; Computer science; Artificial intelligence; Metric (unit); Evaluation of machine translation; Natural language processing; Baseline (sea); Translation (biology); BLEU; Sentence; Language model; Machine learning; Example-based machine translation; Machine translation software usability","authors":[{"name":"Libin Shen","is_ca":false},{"name":"Anoop Sarkar","is_ca":true},{"name":"Franz Josef Och","is_ca":false}],"retraction":null,"screen_n_in":null,"score":{"opus":0.02568892978725502,"gpt":0.3028351473361521,"spread":0.2771462175488971,"validation_status":"score_only:v0-immature-baseline"},"prediction":{"model_version":"metacan-v3-hybrid-931329e0061c","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.002197544,0.00116044,0.001538738,0.001607817,0.000749138,0.001080597,0.001294374,0.001083529,0.003139371],"category_scores_gemma":[0.008533552,0.0003957372,0.0004943817,0.003181815,0.0009679261,0.002307912,0.0008983543,0.001710014,0.002725045],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0006617381,"about_ca_system_score_gemma":0.0009966285,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.001643801,"about_ca_topic_score_gemma":0.002892622,"domain_scores_codex":[0.9977667,0.00115339,0.0001132845,0.0003728516,0.0005090286,0.00008476605],"domain_scores_gemma":[0.9957783,0.002522293,0.0003481506,0.0007077409,0.0005634869,0.00007998724],"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":[0.0001444568,0.0001241621,0.0004900495,0.0005610146,0.00008052823,0.000147986,0.0001350333,0.107424,0.01117808,0.07019607,0.01441087,0.7951078],"study_design_scores_gemma":[0.00007816669,0.0001507471,0.0004279791,0.00003421875,0.00003711125,0.000257708,0.00002683676,0.8296517,0.007926566,0.142722,0.01863948,0.00004756615],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.002488968,0.001599996,0.9936591,0.0001617652,0.0001040527,0.00003100873,0.00006571405,0.0009137604,0.0009755775],"genre_scores_gemma":[0.1501319,0.002715604,0.8389713,0.000262614,0.0008632711,0.0002565981,0.0009915043,0.0003061066,0.005500982],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.003139371,"threshold_uncertainty_score":0.01162189,"prediction_status":"machine_predicted_unvalidated"},"labels":[],"label_agreement":null},{"id":"W1493557574","doi":"10.4324/9781315749129","title":"Routledge Encyclopedia of Translation Technology","year":2014,"lang":"en","type":"book","venue":"","topic":"Natural Language Processing Techniques","field":"Computer Science","cited_by":202,"is_retracted":false,"has_abstract":true,"routes":{"ca_aff":false,"ca_fund":false,"ca_venue":false,"about_ca":true},"ca_institutions":"","funders":"","keywords":"Encyclopedia; Translation (biology); History; Library science; Computer science; Biology","authors":[],"retraction":null,"screen_n_in":null,"score":{"opus":0.008834097611382362,"gpt":0.2447141565008683,"spread":0.2358800588894859,"validation_status":"score_only:v0-immature-baseline"},"prediction":{"model_version":"metacan-v3-hybrid-931329e0061c","candidate_categories":["insufficient_payload"],"consensus_categories":[],"category_scores_codex":[0.001060369,0.001330218,0.001441355,0.004399066,0.001865581,0.01084648,0.002016742,0.00270321,0.303146],"category_scores_gemma":[0.00546538,0.0006905882,0.0007017603,0.01214653,0.002092269,0.009790854,0.003562457,0.00312961,0.2223444],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.003752448,"about_ca_system_score_gemma":0.007038189,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.007887505,"about_ca_topic_score_gemma":0.006685067,"domain_scores_codex":[0.9984906,0.0003159812,0.0002294443,0.0002208632,0.000603882,0.0001392101],"domain_scores_gemma":[0.9979135,0.00074569,0.0001371784,0.0004544512,0.0006251557,0.0001240638],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"not_applicable","study_design_gemma":"not_applicable","study_design_scores_codex":[0.00002582776,0.00001975244,0.0001308198,0.001457614,0.00001105529,0.0002607007,0.001476431,0.0002191344,0.000354717,0.08030535,0.61781,0.2979287],"study_design_scores_gemma":[0.000001687734,0.000002569063,0.00007445336,0.0003782195,0.000001605235,0.00007896615,0.0002212665,0.00003181934,0.00004034197,0.004128258,0.9950362,0.000004621347],"study_design_candidate":"not_applicable","study_design_consensus":"not_applicable","genre_codex":"other","genre_gemma":"other","genre_scores_codex":[0.0005931752,0.09674465,0.006602676,0.01152676,0.007877338,0.0001544648,0.005775874,0.001596251,0.8691288],"genre_scores_gemma":[0.01021449,0.1011569,0.00507023,0.003167061,0.001907382,0.0003582394,0.006258158,0.001541918,0.8703256],"genre_candidate":"other","genre_consensus":"other","teacher_disagreement_score":0.303146,"threshold_uncertainty_score":0.9939777,"prediction_status":"machine_predicted_unvalidated"},"labels":[],"label_agreement":null},{"id":"W2759181158","doi":"10.18653/v1/2023.acl-short","title":"Proceedings of the 61st Annual Meeting of the Association for Computational Linguistics (Volume 2: Short Papers)","year":2023,"lang":"en","type":"paratext","venue":"","topic":"Natural Language Processing Techniques","field":"Computer Science","cited_by":201,"is_retracted":false,"has_abstract":true,"routes":{"ca_aff":false,"ca_fund":true,"ca_venue":false,"about_ca":true},"ca_institutions":"","funders":"Atomic Energy of Canada Limited","keywords":"Volume (thermodynamics); Computational linguistics; Computer science; Association (psychology); Natural language processing; Psychology; Physics","authors":[],"retraction":null,"screen_n_in":null,"score":{"opus":0.0122143289876546,"gpt":0.2760036831266697,"spread":0.2637893541390151,"validation_status":"score_only:v0-immature-baseline"},"prediction":{"model_version":"metacan-v3-hybrid-931329e0061c","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.01224531,0.001165719,0.002668333,0.005588764,0.0027582,0.01212592,0.002686314,0.003729936,0.2617745],"category_scores_gemma":[0.03208539,0.001040149,0.001193451,0.004587137,0.002112811,0.01021868,0.003736968,0.004847599,0.2136264],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.002734267,"about_ca_system_score_gemma":0.005572435,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.004120241,"about_ca_topic_score_gemma":0.005488215,"domain_scores_codex":[0.9950827,0.001589095,0.0007620789,0.0009208426,0.001318627,0.0003267028],"domain_scores_gemma":[0.9701073,0.01214659,0.001173028,0.003007913,0.01039743,0.003167707],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"not_applicable","study_design_gemma":"not_applicable","study_design_scores_codex":[0.00005466466,0.00003656861,0.0004510254,0.0003574498,0.00002815302,0.00005916535,0.0001362274,0.00002526732,0.0003232588,0.00116929,0.9388952,0.05846377],"study_design_scores_gemma":[0.00001624224,0.00001723488,0.0009631094,0.0005457109,0.00001986481,0.0001302732,0.0002466537,0.0001416352,0.0001132297,0.002124222,0.9956631,0.00001884098],"study_design_candidate":"not_applicable","study_design_consensus":"not_applicable","genre_codex":"editorial","genre_gemma":"other","genre_scores_codex":[0.007923364,0.1487252,0.03583389,0.1830549,0.3668381,0.00100055,0.02258789,0.006088659,0.2279474],"genre_scores_gemma":[0.02362778,0.08117965,0.03749578,0.03801403,0.07413077,0.001778485,0.04154314,0.0075707,0.6946597],"genre_candidate":"other","genre_consensus":null,"teacher_disagreement_score":0.2617745,"threshold_uncertainty_score":0.8757228,"prediction_status":"machine_predicted_unvalidated"},"labels":[],"label_agreement":null}]}