{"meta":{"page":1,"per_page":50,"max_per_page":100,"total":22,"total_is_capped":false,"direct_labels_cover":0,"predictions_cover":22,"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":"5250998b0ea9","filters":{"venue":"Meeting of the Association for Computational Linguistics"}},"results":[{"id":"W2115791615","doi":"","title":"Deep Learning for NLP (without Magic)","year":2012,"lang":"en","type":"article","venue":"Meeting of the Association for Computational Linguistics","topic":"Topic Modeling","field":"Computer Science","cited_by":132,"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; Artificial intelligence; Feature engineering; Deep learning; Artificial neural network; Paraphrase; Machine learning; Sentiment analysis; Natural language processing; Feature (linguistics); Focus (optics); MAGIC (telescope); Language model","authors":[{"name":"Richard Socher","is_ca":false},{"name":"Christopher D. Manning","is_ca":true}],"retraction":null,"screen_n_in":null,"score":{"opus":0.02272081690193958,"gpt":0.2786095360606215,"spread":0.2558887191586819,"validation_status":"score_only:v0-immature-baseline"},"prediction":{"model_version":"codex-gemma-dda1882f352a","candidate_categories":["metaresearch"],"consensus_categories":[],"category_scores_codex":[0.001386031,0.00008897814,0.0001465913,0.00004466897,0.0002715792,0.00005278375,0.000402146,0.00006137528,5.789303e-7],"category_scores_gemma":[0.01366101,0.00007997827,0.0001360499,0.0001205166,0.00001115577,0.00005187264,0.0001186465,0.00009763151,0.000003059626],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.000156419,"about_ca_system_score_gemma":0.00005087542,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.000006018312,"about_ca_topic_score_gemma":0.000001318756,"domain_scores_codex":[0.9988171,0.00007260493,0.0003410196,0.0001474626,0.0003618822,0.0002598563],"domain_scores_gemma":[0.996572,0.001635076,0.0005976237,0.0001409311,0.00101508,0.00003930101],"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":[0.000004978131,0.00004159337,0.08693471,0.00004713894,0.0000416298,1.527839e-8,0.0007317339,0.6599267,0.00001894471,0.2508814,0.0002900635,0.001081114],"study_design_scores_gemma":[0.0003144084,0.00001799103,0.005148931,0.00002381664,0.0000259264,3.424852e-7,0.00001970639,0.9410582,0.0001378555,0.03734311,0.01581066,0.00009906648],"study_design_candidate":"simulation_or_modeling","study_design_consensus":"simulation_or_modeling","genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.00284011,0.00006412678,0.9922226,0.0003184541,0.001948096,0.0002811218,0.000005569278,0.00006671975,0.002253204],"genre_scores_gemma":[0.6895633,4.095321e-7,0.3092426,0.00008589859,0.000628208,0.00001066278,0.00001033502,0.00000916903,0.0004493714],"genre_candidate":"methods","genre_consensus":null,"teacher_disagreement_score":0.6867232,"threshold_uncertainty_score":0.9946473,"prediction_status":"machine_predicted_unvalidated"},"labels":[],"label_agreement":null},{"id":"W2120158336","doi":"","title":"Cross Lingual Adaptation: An Experiment on Sentiment Classifications","year":2010,"lang":"en","type":"article","venue":"Meeting of the Association for Computational Linguistics","topic":"Sentiment Analysis and Opinion Mining","field":"Computer Science","cited_by":87,"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":"Computer science; Machine translation; Natural language processing; Task (project management); Artificial intelligence; Representation (politics); Adaptation (eye); Key (lock); Sentiment analysis; Natural language; Translation (biology); Parallel corpora; Noise (video); Machine learning","authors":[{"name":"Bin Wei","is_ca":false},{"name":"Christopher Pal","is_ca":true}],"retraction":null,"screen_n_in":null,"score":{"opus":0.03508279994780869,"gpt":0.3290336885755691,"spread":0.2939508886277604,"validation_status":"score_only:v0-immature-baseline"},"prediction":{"model_version":"codex-gemma-dda1882f352a","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.000953677,0.0001142688,0.0001446995,0.00008717085,0.000476904,0.0001998945,0.0005706583,0.00007237362,0.000006991809],"category_scores_gemma":[0.002722565,0.000100698,0.0001619396,0.0002509195,0.00003070436,0.00005786775,0.00008577656,0.0001422931,0.000007654415],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.00009750696,"about_ca_system_score_gemma":0.0001218683,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.00001017738,"about_ca_topic_score_gemma":0.000006983507,"domain_scores_codex":[0.9983418,0.00007446346,0.00047755,0.0002771562,0.000659505,0.0001695121],"domain_scores_gemma":[0.9967967,0.0007682408,0.0007480322,0.0003326317,0.001301706,0.00005263816],"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":[0.00001099779,0.000320284,0.01153902,0.000009576335,0.0001002508,1.429828e-7,0.001673762,0.4694085,0.0008166462,0.5146855,0.0005847968,0.000850466],"study_design_scores_gemma":[0.0004999574,0.00007538285,0.01293974,0.00002421486,0.00003163579,3.079213e-7,0.0001021594,0.9576163,0.006624462,0.009493971,0.01242116,0.000170668],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.4620585,0.00005701213,0.4946204,0.003455406,0.01879621,0.001440325,0.0001503809,0.000409696,0.0190121],"genre_scores_gemma":[0.8577216,3.676394e-7,0.141042,0.0001220036,0.0004380752,0.00001723681,0.00006025962,0.000009478051,0.0005889419],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.5051915,"threshold_uncertainty_score":0.4106347,"prediction_status":"machine_predicted_unvalidated"},"labels":[],"label_agreement":null},{"id":"W2161427841","doi":"","title":"Summarizing Emails with Conversational Cohesion and Subjectivity","year":2008,"lang":"en","type":"article","venue":"Meeting of the Association for Computational Linguistics","topic":"Advanced Text Analysis Techniques","field":"Computer Science","cited_by":66,"is_retracted":false,"has_abstract":true,"routes":{"ca_aff":true,"ca_fund":false,"ca_venue":false,"about_ca":false},"ca_institutions":"University of British Columbia","funders":"","keywords":"Automatic summarization; Cohesion (chemistry); Computer science; PageRank; Cosine similarity; Natural language processing; Artificial intelligence; Graph; Sentence; Empirical research; Information retrieval; Subjectivity; Theoretical computer science; Pattern recognition (psychology); Mathematics","authors":[{"name":"Giuseppe Carenini","is_ca":true},{"name":"Raymond T. Ng","is_ca":true},{"name":"Xiaodong Zhou","is_ca":true}],"retraction":null,"screen_n_in":null,"score":{"opus":0.01336756495223629,"gpt":0.2436825173727591,"spread":0.2303149524205229,"validation_status":"score_only:v0-immature-baseline"},"prediction":{"model_version":"codex-gemma-dda1882f352a","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.0005475976,0.00008702647,0.0001557552,0.00006476924,0.0003661705,0.00002804876,0.0002327875,0.00004476954,2.688988e-7],"category_scores_gemma":[0.00470842,0.00007067721,0.00006115325,0.0002436779,0.00005630504,0.00006903361,0.0001041411,0.00007745086,4.074625e-7],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0001441364,"about_ca_system_score_gemma":0.0001117305,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.00001600998,"about_ca_topic_score_gemma":0.000007850972,"domain_scores_codex":[0.998931,0.00006599862,0.0002368311,0.000188357,0.0004577815,0.0001200251],"domain_scores_gemma":[0.9962969,0.001856415,0.0005642357,0.000132519,0.001124882,0.00002501478],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"observational","study_design_gemma":"simulation_or_modeling","study_design_scores_codex":[0.00003608725,0.0001057225,0.6116477,0.00005273688,0.0001581404,0.000001944237,0.0006961052,0.07089256,0.0002573885,0.3148366,0.00085511,0.0004598935],"study_design_scores_gemma":[0.001315175,0.0002075095,0.2991844,0.0001450173,0.0001055637,0.00001835701,0.00003321941,0.5126092,0.006137724,0.1761356,0.003662907,0.0004452879],"study_design_candidate":"observational","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.07226845,0.00003212084,0.9253579,0.0004662924,0.0002111419,0.0002940832,0.00001888923,0.0001294875,0.001221631],"genre_scores_gemma":[0.7667304,0.000002116247,0.2330418,0.00005520412,0.00006129601,0.000005775817,0.000008838416,0.000005578057,0.00008905098],"genre_candidate":"methods","genre_consensus":null,"teacher_disagreement_score":0.6944619,"threshold_uncertainty_score":0.5636758,"prediction_status":"machine_predicted_unvalidated"},"labels":[],"label_agreement":null},{"id":"W2124618303","doi":"","title":"Substring-Based Transliteration","year":2007,"lang":"en","type":"article","venue":"Meeting of the Association for Computational Linguistics","topic":"Natural Language Processing Techniques","field":"Computer Science","cited_by":61,"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":"Transliteration; Substring; Computer science; Margin (machine learning); Artificial intelligence; Word (group theory); Natural language processing; Machine translation; Speech recognition; Programming language; Machine learning; Data structure; Mathematics","authors":[{"name":"Tarek Sherif","is_ca":true},{"name":"Grzegorz Kondrak","is_ca":true}],"retraction":null,"screen_n_in":null,"score":{"opus":0.01112075728909654,"gpt":0.2755006221297466,"spread":0.2643798648406501,"validation_status":"score_only:v0-immature-baseline"},"prediction":{"model_version":"codex-gemma-dda1882f352a","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.001559742,0.00007827007,0.0001029536,0.00008023988,0.0001817752,0.00006934439,0.0004611382,0.000069058,2.966945e-7],"category_scores_gemma":[0.005592148,0.00006698395,0.00009981616,0.0003080182,0.00001507392,0.000039277,0.00003635957,0.00009306314,5.667156e-7],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0001647872,"about_ca_system_score_gemma":0.0000920859,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.000006651142,"about_ca_topic_score_gemma":0.000006094109,"domain_scores_codex":[0.9988784,0.0000380027,0.0003499077,0.0001443305,0.0004333016,0.0001560965],"domain_scores_gemma":[0.9968214,0.00124031,0.0005063826,0.0001406427,0.00126878,0.00002246633],"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":[0.00003124128,0.0001176552,0.02060572,0.0001353034,0.00004419144,8.324939e-7,0.0006276225,0.1231834,0.0009057717,0.8503111,0.0008605451,0.003176628],"study_design_scores_gemma":[0.0008609492,0.00007698693,0.00763366,0.0001894939,0.00004153016,9.210525e-7,0.000009947464,0.5274749,0.06982344,0.3864295,0.007157406,0.0003012496],"study_design_candidate":"theoretical_or_conceptual","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.001962826,0.00008958159,0.995899,0.0003936715,0.0006113058,0.0001824885,0.00001247981,0.0001530553,0.000695547],"genre_scores_gemma":[0.5837101,7.615142e-8,0.4160012,0.0001167449,0.0001091209,0.00000223083,0.00001129261,0.000004575466,0.00004457874],"genre_candidate":"methods","genre_consensus":null,"teacher_disagreement_score":0.5817474,"threshold_uncertainty_score":0.6694728,"prediction_status":"machine_predicted_unvalidated"},"labels":[],"label_agreement":null},{"id":"W2096565906","doi":"","title":"Alignment-Based Discriminative String Similarity","year":2007,"lang":"en","type":"article","venue":"Meeting of the Association for Computational Linguistics","topic":"Natural Language Processing Techniques","field":"Computer Science","cited_by":54,"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":"Discriminative model; Coreference; Artificial intelligence; Computer science; Substring; String metric; Similarity (geometry); Character (mathematics); Natural language processing; Longest common subsequence problem; String (physics); Word (group theory); Pattern recognition (psychology); Transliteration; Heuristic; String searching algorithm; Mathematics; Pattern matching; Resolution (logic); Algorithm; Set (abstract data type)","authors":[{"name":"Shane Bergsma","is_ca":true},{"name":"Grzegorz Kondrak","is_ca":true}],"retraction":null,"screen_n_in":null,"score":{"opus":0.01628738332415331,"gpt":0.2953723765151392,"spread":0.2790849931909859,"validation_status":"score_only:v0-immature-baseline"},"prediction":{"model_version":"codex-gemma-dda1882f352a","candidate_categories":["metaresearch"],"consensus_categories":[],"category_scores_codex":[0.001767842,0.0000954509,0.0001281573,0.00007461118,0.000236727,0.00006207456,0.000590612,0.00006804157,3.179561e-7],"category_scores_gemma":[0.01115894,0.00008052967,0.0001038899,0.0002699594,0.00002553938,0.00004011754,0.0001401127,0.0001189502,4.31109e-7],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0003126034,"about_ca_system_score_gemma":0.0001006165,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.00001443478,"about_ca_topic_score_gemma":0.000006401025,"domain_scores_codex":[0.9986812,0.00005412094,0.0003449237,0.0001759115,0.0005483596,0.0001954224],"domain_scores_gemma":[0.9960378,0.001865358,0.0006803975,0.0001589809,0.001228894,0.00002860524],"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":[0.00001793994,0.0001201039,0.01440984,0.00009839574,0.00005125125,9.584106e-7,0.000540001,0.03079733,0.0003517079,0.9508581,0.0009003515,0.001854038],"study_design_scores_gemma":[0.0006834052,0.00007447046,0.00689848,0.0002026356,0.00005455118,5.160374e-7,0.00004679305,0.2728774,0.0717092,0.6446496,0.002502135,0.0003008987],"study_design_candidate":"theoretical_or_conceptual","study_design_consensus":"theoretical_or_conceptual","genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.001355758,0.00005870457,0.9961205,0.000482007,0.0006019163,0.0002101979,0.00002537581,0.0001409256,0.001004591],"genre_scores_gemma":[0.5739122,9.164423e-8,0.425802,0.0001284454,0.00009099129,0.000002477214,0.000008945046,0.000005015394,0.0000498364],"genre_candidate":"methods","genre_consensus":null,"teacher_disagreement_score":0.5725564,"threshold_uncertainty_score":0.9971705,"prediction_status":"machine_predicted_unvalidated"},"labels":[],"label_agreement":null},{"id":"W2097120204","doi":"","title":"Towards Robust Abstractive Multi-Document Summarization: A Caseframe Analysis of Centrality and Domain","year":2013,"lang":"en","type":"article","venue":"Meeting of the Association for Computational Linguistics","topic":"Topic Modeling","field":"Computer Science","cited_by":29,"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":"Automatic summarization; Centrality; Computer science; Redundancy (engineering); Sentence; Natural language processing; Information retrieval; Domain (mathematical analysis); Artificial intelligence; Abstraction; Multi-document summarization","authors":[{"name":"Jackie Chi Kit Cheung","is_ca":true},{"name":"Gerald Penn","is_ca":true}],"retraction":null,"screen_n_in":null,"score":{"opus":0.02083360009022558,"gpt":0.268047448719079,"spread":0.2472138486288535,"validation_status":"score_only:v0-immature-baseline"},"prediction":{"model_version":"codex-gemma-dda1882f352a","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.0007165367,0.00008428045,0.0002188145,0.00009425075,0.0001245886,0.00006578765,0.0002668058,0.00005765325,0.000002280509],"category_scores_gemma":[0.003441396,0.00007498434,0.0001235562,0.000380584,0.0000256593,0.00005576037,0.0001265226,0.00006738471,4.100677e-7],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0001600274,"about_ca_system_score_gemma":0.00008302229,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.0003277853,"about_ca_topic_score_gemma":0.00002395716,"domain_scores_codex":[0.9987148,0.00009074327,0.0004642716,0.0001922088,0.0004043157,0.0001336741],"domain_scores_gemma":[0.9967899,0.0007865981,0.0007991745,0.0001690555,0.001420578,0.00003464952],"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":[0.000002512956,0.00005545706,0.04691549,0.00003224802,0.000370927,9.230399e-8,0.0008279551,0.8622157,0.00002814869,0.08913369,0.00004967461,0.0003681038],"study_design_scores_gemma":[0.000257538,0.0000112459,0.1402108,0.00001840662,0.0001366933,1.302402e-7,0.00004619604,0.8291977,0.0001235714,0.02982892,0.00009658014,0.00007215013],"study_design_candidate":"simulation_or_modeling","study_design_consensus":"simulation_or_modeling","genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.07560828,0.00002220825,0.9226575,0.0005519021,0.0003673323,0.0003216251,0.00005988277,0.00002334051,0.0003879216],"genre_scores_gemma":[0.6911564,8.515854e-7,0.3086889,0.00003868679,0.00004097073,0.000006866563,0.00001950623,0.000003270019,0.00004454197],"genre_candidate":"methods","genre_consensus":null,"teacher_disagreement_score":0.6155481,"threshold_uncertainty_score":0.4119921,"prediction_status":"machine_predicted_unvalidated"},"labels":[],"label_agreement":null},{"id":"W2110750141","doi":"","title":"Automatic detection of deception in child-produced speech using syntactic complexity features","year":2013,"lang":"en","type":"article","venue":"Meeting of the Association for Computational Linguistics","topic":"Deception detection and forensic psychology","field":"Psychology","cited_by":27,"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":"Utterance; Deception; Sentence; Support vector machine; Computer science; Set (abstract data type); Artificial intelligence; Random forest; Measure (data warehouse); Natural language processing; Speech recognition; Pattern recognition (psychology); Machine learning; Psychology; Data mining; Social psychology","authors":[{"name":"Maria Yancheva","is_ca":true},{"name":"Frank Rudzicz","is_ca":true}],"retraction":null,"screen_n_in":null,"score":{"opus":0.03243081732255438,"gpt":0.3234808616065863,"spread":0.291050044284032,"validation_status":"score_only:v0-immature-baseline"},"prediction":{"model_version":"codex-gemma-dda1882f352a","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.0007643191,0.0001200487,0.0002600997,0.0002045621,0.00014569,0.00001804265,0.0001533217,0.0001507796,0.00007809643],"category_scores_gemma":[0.004876104,0.0001133603,0.000144234,0.0003892995,0.00007449533,0.00002821908,0.00003119618,0.000184359,0.0000176088],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.000251493,"about_ca_system_score_gemma":0.00002964392,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.0003854388,"about_ca_topic_score_gemma":0.00006906747,"domain_scores_codex":[0.9983708,0.0002461965,0.0006561288,0.0002152927,0.0003201128,0.0001915255],"domain_scores_gemma":[0.9971058,0.0006224396,0.00112255,0.0001858206,0.0009360791,0.00002724575],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"observational","study_design_gemma":"observational","study_design_scores_codex":[0.0009778397,0.00451922,0.450848,0.001324682,0.002258034,0.000003843327,0.01369118,0.2577409,0.05470102,0.1234475,0.01193654,0.07855128],"study_design_scores_gemma":[0.001791282,0.0001362923,0.8587577,0.0001611623,0.0001052659,0.0000237106,0.0004831126,0.08702328,0.003705894,0.04707916,0.0004890181,0.0002441635],"study_design_candidate":"observational","study_design_consensus":"observational","genre_codex":"empirical","genre_gemma":"empirical","genre_scores_codex":[0.9874063,0.00002908127,0.003365439,0.0002233974,0.004065541,0.0006868414,0.00003181452,0.00005146808,0.004140131],"genre_scores_gemma":[0.9927102,5.488591e-7,0.006734697,0.00007188527,0.0002594093,0.00002252249,0.00001965745,0.00001766552,0.0001633927],"genre_candidate":"empirical","genre_consensus":"empirical","teacher_disagreement_score":0.4079097,"threshold_uncertainty_score":0.5837504,"prediction_status":"machine_predicted_unvalidated"},"labels":[],"label_agreement":null},{"id":"W2166202273","doi":"","title":"Bootstrapping a Stochastic Transducer for Arabic-English Transliteration Extraction","year":2007,"lang":"en","type":"article","venue":"Meeting of the Association for Computational Linguistics","topic":"Natural Language Processing Techniques","field":"Computer Science","cited_by":24,"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":"Bootstrapping (finance); Computer science; Transducer; Scripting language; Artificial intelligence; Natural language processing; Metric (unit); Transliteration; Task (project management); Speech recognition; Similarity (geometry); Arabic; Pattern recognition (psychology); Programming language; Linguistics; Acoustics; Engineering; Mathematics","authors":[{"name":"Tarek Sherif","is_ca":true},{"name":"Grzegorz Kondrak","is_ca":true}],"retraction":null,"screen_n_in":null,"score":{"opus":0.01409818308605324,"gpt":0.2948093699806726,"spread":0.2807111868946193,"validation_status":"score_only:v0-immature-baseline"},"prediction":{"model_version":"codex-gemma-dda1882f352a","candidate_categories":["metaresearch"],"consensus_categories":[],"category_scores_codex":[0.001843879,0.0001185635,0.0001574573,0.0001087576,0.0002947205,0.0001090874,0.0003834029,0.0001103324,3.237761e-7],"category_scores_gemma":[0.008877081,0.0001077997,0.0001691128,0.0002723123,0.00001818872,0.0001109761,0.00002141879,0.0001322181,2.079145e-7],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0002146391,"about_ca_system_score_gemma":0.00009423139,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.000005333234,"about_ca_topic_score_gemma":0.000007139969,"domain_scores_codex":[0.9985852,0.00004039236,0.0005023945,0.000225509,0.0004214443,0.0002250903],"domain_scores_gemma":[0.994791,0.00201069,0.0006373667,0.0001360332,0.002394087,0.00003087454],"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":[0.0001580797,0.0002601143,0.001001163,0.0004552295,0.0001784021,5.42667e-7,0.007071889,0.2793137,0.004677092,0.6970106,0.001073092,0.008800037],"study_design_scores_gemma":[0.001140067,0.0001226485,0.001334392,0.0002758622,0.0001040861,0.000001910839,0.00006865757,0.6004904,0.01448079,0.3775532,0.004064955,0.0003630483],"study_design_candidate":"theoretical_or_conceptual","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.001292436,0.0001137159,0.995942,0.0002525005,0.001359611,0.0005494583,0.00003563123,0.0001889923,0.0002656652],"genre_scores_gemma":[0.6566775,2.270905e-7,0.3427181,0.0000710173,0.0004186222,0.00001735127,0.00002631695,0.000009989126,0.00006089823],"genre_candidate":"methods","genre_consensus":null,"teacher_disagreement_score":0.6553851,"threshold_uncertainty_score":0.9994715,"prediction_status":"machine_predicted_unvalidated"},"labels":[],"label_agreement":null},{"id":"W4531093","doi":"","title":"Extraction of Disease-Treatment Semantic Relations from Biomedical Sentences","year":2010,"lang":"en","type":"article","venue":"Meeting of the Association for Computational Linguistics","topic":"Biomedical Text Mining and Ontologies","field":"Biochemistry, Genetics and Molecular Biology","cited_by":23,"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":"Relationship extraction; Relation (database); Semantic relation; Measure (data warehouse); Computer science; Natural language processing; Focus (optics); Artificial intelligence; Semantics (computer science); Information retrieval; Information extraction; Data mining; Medicine; Programming language","authors":[{"name":"Oana Frunza","is_ca":true},{"name":"Diana Inkpen","is_ca":true}],"retraction":null,"screen_n_in":null,"score":{"opus":0.01283156771914793,"gpt":0.2901257257106199,"spread":0.2772941579914719,"validation_status":"score_only:v0-immature-baseline"},"prediction":{"model_version":"codex-gemma-dda1882f352a","candidate_categories":["metaresearch"],"consensus_categories":[],"category_scores_codex":[0.000222129,0.00006168886,0.00009863279,0.00002261564,0.00009662801,0.000006420848,0.0001002647,0.0001001221,0.000003110756],"category_scores_gemma":[0.009224476,0.00004705204,0.0001069998,0.00005967411,0.00008603546,7.028161e-7,0.00002861608,0.00005451886,9.258395e-7],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.00002168745,"about_ca_system_score_gemma":0.0001175416,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.00005089888,"about_ca_topic_score_gemma":0.00002692829,"domain_scores_codex":[0.9993033,0.00003795448,0.0002497548,0.0001178598,0.0002118608,0.00007930899],"domain_scores_gemma":[0.9985871,0.0005057341,0.0003849569,0.00009916763,0.0003896368,0.00003334764],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"observational","study_design_gemma":"observational","study_design_scores_codex":[0.000324681,0.001178864,0.7964888,0.0001120162,0.0008811395,7.507383e-7,0.0004954164,0.02307677,0.1544357,0.008617773,0.009253093,0.005134956],"study_design_scores_gemma":[0.002513095,0.0006772339,0.6330354,0.0002006814,0.0006411771,0.000001411847,0.0002980308,0.05860434,0.03172022,0.05106872,0.22079,0.0004497399],"study_design_candidate":"observational","study_design_consensus":"observational","genre_codex":"empirical","genre_gemma":"empirical","genre_scores_codex":[0.9875615,0.00008435921,0.008876423,0.0007207633,0.001789853,0.0001635303,0.0004209989,0.00001428709,0.0003683304],"genre_scores_gemma":[0.9793823,0.000004108637,0.01952213,0.00001899902,0.0004609326,0.000006162283,0.000331609,0.00000570901,0.000268101],"genre_candidate":"empirical","genre_consensus":"empirical","teacher_disagreement_score":0.2115369,"threshold_uncertainty_score":0.9991212,"prediction_status":"machine_predicted_unvalidated"},"labels":[],"label_agreement":null},{"id":"W2099233844","doi":"","title":"Even the Abstract have Color: Consensus in Word-Colour Associations","year":2011,"lang":"en","type":"article","venue":"Meeting of the Association for Computational Linguistics","topic":"Categorization, perception, and language","field":"Psychology","cited_by":16,"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":"Lexicon; Computer science; Word Association; Focus (optics); Component (thermodynamics); Crowdsourcing; Word (group theory); Resource (disambiguation); Key (lock); Product (mathematics); Association (psychology); Visualization; Semantics (computer science); Natural language processing; Coherence (philosophical gambling strategy); Quality (philosophy); Artificial intelligence; Linguistics; World Wide Web; Psychology; Mathematics","authors":[{"name":"Saif M. Mohammad","is_ca":true}],"retraction":null,"screen_n_in":null,"score":{"opus":0.0455092320359143,"gpt":0.3172408818883953,"spread":0.271731649852481,"validation_status":"score_only:v0-immature-baseline"},"prediction":{"model_version":"codex-gemma-dda1882f352a","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.001502092,0.0001199405,0.0001926402,0.00008159353,0.0003185756,0.00002475186,0.0002621796,0.0001473179,0.00009117548],"category_scores_gemma":[0.007912361,0.00009551436,0.000149591,0.0002379361,0.00005829966,0.00001162036,0.00003513597,0.0001824043,0.0000213973],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0003132924,"about_ca_system_score_gemma":0.0001295204,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.0007931062,"about_ca_topic_score_gemma":0.0008472526,"domain_scores_codex":[0.9984413,0.0001991855,0.0006013266,0.0001844815,0.0003411498,0.0002325917],"domain_scores_gemma":[0.9956473,0.002221782,0.0009439074,0.0001782239,0.0009821827,0.00002656704],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"observational","study_design_gemma":"observational","study_design_scores_codex":[0.000110926,0.0005455359,0.7695939,0.00002930514,0.0003235454,0.00000158064,0.03389458,0.02828078,0.00004431502,0.1521519,0.01471799,0.0003057003],"study_design_scores_gemma":[0.0006115649,0.00002731178,0.9663225,0.00002117397,0.00007999379,8.511524e-7,0.001568838,0.002520736,0.00002669017,0.02489579,0.00379775,0.000126852],"study_design_candidate":"observational","study_design_consensus":"observational","genre_codex":"empirical","genre_gemma":"empirical","genre_scores_codex":[0.8539451,0.0001307819,0.0009540445,0.001319299,0.005850151,0.001728782,0.0008925904,0.0001273119,0.1350519],"genre_scores_gemma":[0.9951062,0.000001138254,0.001416044,0.0001807991,0.0004105826,0.00004046365,0.0001264875,0.00002127541,0.002696968],"genre_candidate":"empirical","genre_consensus":"empirical","teacher_disagreement_score":0.1967286,"threshold_uncertainty_score":0.9472407,"prediction_status":"machine_predicted_unvalidated"},"labels":[],"label_agreement":null},{"id":"W2096960653","doi":"","title":"Selecting Query Term Alternations for Web Search by Exploiting Query Contexts","year":2008,"lang":"en","type":"article","venue":"Meeting of the Association for Computational Linguistics","topic":"Algorithms and Data Compression","field":"Computer Science","cited_by":13,"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":"Query expansion; Computer science; Bigram; Web query classification; Web search query; Information retrieval; Query optimization; Query language; Sargable; Selection (genetic algorithm); Term (time); RDF query language; Context (archaeology); Search engine; Word (group theory); Natural language processing; Artificial intelligence","authors":[{"name":"Guihong Cao","is_ca":true},{"name":"Stephen Robertson","is_ca":false},{"name":"Jian‐Yun Nie","is_ca":true}],"retraction":null,"screen_n_in":null,"score":{"opus":0.02470986278797762,"gpt":0.2848725438985707,"spread":0.2601626811105931,"validation_status":"score_only:v0-immature-baseline"},"prediction":{"model_version":"codex-gemma-dda1882f352a","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.0009835619,0.000125909,0.0001948076,0.00007691269,0.000940868,0.0000880641,0.0006080313,0.00007112412,5.762763e-7],"category_scores_gemma":[0.007220434,0.000115733,0.0001488135,0.000226584,0.000029502,0.0001022479,0.0002062924,0.0001301607,0.000001503921],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0002098003,"about_ca_system_score_gemma":0.0002481011,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.00002955192,"about_ca_topic_score_gemma":0.000003477346,"domain_scores_codex":[0.9982846,0.00009652114,0.000493121,0.0002763001,0.0005761903,0.0002732062],"domain_scores_gemma":[0.9942311,0.003016955,0.0006672537,0.0001924795,0.001845116,0.00004707024],"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":[0.00009834574,0.0009309679,0.1645883,0.000487604,0.0006312722,0.000003494818,0.005147472,0.3884861,0.008253797,0.2907028,0.1309308,0.009739125],"study_design_scores_gemma":[0.001000111,0.0000646183,0.004994031,0.000139465,0.00002404389,0.000003800428,0.00003610508,0.9676545,0.002917867,0.01048574,0.01244371,0.0002360264],"study_design_candidate":"simulation_or_modeling","study_design_consensus":"simulation_or_modeling","genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.04061704,0.0001078574,0.9545274,0.0007071448,0.001796078,0.0006706764,0.0004664855,0.0001501319,0.0009571853],"genre_scores_gemma":[0.8465438,0.000004053579,0.1521304,0.0001542564,0.0005445436,0.00003483465,0.0001442609,0.00001759295,0.0004261856],"genre_candidate":"methods","genre_consensus":null,"teacher_disagreement_score":0.8059268,"threshold_uncertainty_score":0.8644056,"prediction_status":"machine_predicted_unvalidated"},"labels":[],"label_agreement":null},{"id":"W2099242427","doi":"","title":"Entity-Based Local Coherence Modelling Using Topological Fields","year":2010,"lang":"en","type":"article","venue":"Meeting of the Association for Computational Linguistics","topic":"Natural Language Processing Techniques","field":"Computer Science","cited_by":10,"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":"Coherence (philosophical gambling strategy); Computer science; Sentence; Component (thermodynamics); Topology (electrical circuits); Grid; Natural language; Natural language processing; Field (mathematics); Artificial intelligence; Language model; Theoretical computer science; Mathematics; Physics; Pure mathematics","authors":[{"name":"Jackie Chi Kit Cheung","is_ca":true},{"name":"Gerald Penn","is_ca":true}],"retraction":null,"screen_n_in":null,"score":{"opus":0.02298369617306752,"gpt":0.2908722400383604,"spread":0.2678885438652929,"validation_status":"score_only:v0-immature-baseline"},"prediction":{"model_version":"codex-gemma-dda1882f352a","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.0007391617,0.00008863433,0.0001304608,0.00004854589,0.0002479958,0.00008139574,0.0006352109,0.0001349428,0.000001274806],"category_scores_gemma":[0.004431317,0.00007328361,0.0001000261,0.0001980973,0.00004875143,0.00003661312,0.0001278054,0.00025926,4.622486e-7],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.00008819831,"about_ca_system_score_gemma":0.0001593948,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.00004335261,"about_ca_topic_score_gemma":0.000009020398,"domain_scores_codex":[0.9989322,0.00005208386,0.000285987,0.0001793004,0.0003922314,0.0001581449],"domain_scores_gemma":[0.9972548,0.0009648943,0.0004705498,0.0001719638,0.001111164,0.00002660669],"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":[0.000004110468,0.00003511839,0.002199406,0.00002726265,0.000009978698,2.888772e-7,0.00005935924,0.7465371,0.0002566007,0.2505423,0.00007948154,0.0002489723],"study_design_scores_gemma":[0.0001034904,0.00001369249,0.00004334348,0.00002722996,0.00001020242,4.512054e-7,0.000002235628,0.7718817,0.003888295,0.2236966,0.0002615603,0.00007117189],"study_design_candidate":"simulation_or_modeling","study_design_consensus":"simulation_or_modeling","genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.007265297,0.00002492652,0.9909387,0.0002767312,0.0009672953,0.000136676,0.00001077659,0.0001123025,0.0002672912],"genre_scores_gemma":[0.524128,5.526093e-8,0.4756452,0.00009700775,0.0001008023,0.000002254769,0.000003573273,0.000003430804,0.00001961925],"genre_candidate":"methods","genre_consensus":null,"teacher_disagreement_score":0.5168627,"threshold_uncertainty_score":0.5305021,"prediction_status":"machine_predicted_unvalidated"},"labels":[],"label_agreement":null},{"id":"W2146546639","doi":"","title":"Learning Bigrams from Unigrams","year":2008,"lang":"en","type":"article","venue":"Meeting of the Association for Computational Linguistics","topic":"Topic Modeling","field":"Computer Science","cited_by":9,"is_retracted":false,"has_abstract":true,"routes":{"ca_aff":true,"ca_fund":false,"ca_venue":false,"about_ca":false},"ca_institutions":"McGill University","funders":"","keywords":"Bigram; Perplexity; Computer science; Language model; Natural language processing; Artificial intelligence; Word (group theory); Trigram; Oracle; Set (abstract data type); Speech recognition; Mathematics; Programming language","authors":[{"name":"Xiaojin Zhu","is_ca":false},{"name":"Andrew B. Goldberg","is_ca":false},{"name":"Michael Rabbat","is_ca":true},{"name":"Robert Nowak","is_ca":false}],"retraction":null,"screen_n_in":null,"score":{"opus":0.02239389677370867,"gpt":0.2445550466316511,"spread":0.2221611498579424,"validation_status":"score_only:v0-immature-baseline"},"prediction":{"model_version":"codex-gemma-dda1882f352a","candidate_categories":["metaresearch"],"consensus_categories":[],"category_scores_codex":[0.0004997115,0.0000860983,0.0001456079,0.00004260376,0.0003854272,0.00003823262,0.0005293936,0.00006326164,6.484752e-7],"category_scores_gemma":[0.008465158,0.00007786802,0.0001219603,0.000220848,0.00002130962,0.00002987542,0.0001419479,0.0001377924,0.000003884287],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0001264315,"about_ca_system_score_gemma":0.0001070811,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.0001042024,"about_ca_topic_score_gemma":0.000005143499,"domain_scores_codex":[0.9987119,0.00009312051,0.0003462097,0.0001954529,0.0004927404,0.000160614],"domain_scores_gemma":[0.9969376,0.001429117,0.0005488979,0.0001808242,0.0008743603,0.00002917531],"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":[0.000003377751,0.00003773203,0.08573426,0.000009505736,0.0000509084,4.885418e-7,0.001133965,0.8445374,0.0000396203,0.0671602,0.0004813809,0.0008112015],"study_design_scores_gemma":[0.0003314731,0.00002674421,0.01265597,0.00003600351,0.00001507661,6.524019e-7,0.00002167378,0.9194275,0.0002571492,0.05800137,0.009110992,0.0001154303],"study_design_candidate":"simulation_or_modeling","study_design_consensus":"simulation_or_modeling","genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.1152335,0.00005555198,0.8779747,0.0006009102,0.002156079,0.0002231491,0.00002382695,0.0001749083,0.003557381],"genre_scores_gemma":[0.8257209,0.000001755787,0.1734126,0.00008479365,0.0003273247,0.000003964496,0.00002049721,0.00000770738,0.000420408],"genre_candidate":"methods","genre_consensus":null,"teacher_disagreement_score":0.7104874,"threshold_uncertainty_score":0.999887,"prediction_status":"machine_predicted_unvalidated"},"labels":[],"label_agreement":null},{"id":"W2250669704","doi":"","title":"Probabilistic Domain Modelling With Contextualized Distributional Semantic Vectors","year":2013,"lang":"en","type":"article","venue":"Meeting of the Association for Computational Linguistics","topic":"Topic Modeling","field":"Computer Science","cited_by":8,"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":"Automatic summarization; Computer science; Distributional semantics; Artificial intelligence; Probabilistic logic; Natural language processing; Domain (mathematical analysis); Semantics (computer science); Generative grammar; Generative model; Hidden Markov model; Semantic space; Machine learning; Semantic similarity; Mathematics","authors":[{"name":"Jackie Chi Kit Cheung","is_ca":true},{"name":"Gerald Penn","is_ca":true}],"retraction":null,"screen_n_in":null,"score":{"opus":0.01520811403659151,"gpt":0.2314686742057899,"spread":0.2162605601691984,"validation_status":"score_only:v0-immature-baseline"},"prediction":{"model_version":"codex-gemma-dda1882f352a","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.0006579137,0.0001282429,0.0002020068,0.00004536373,0.0002767185,0.0001113458,0.0004845522,0.00006021373,0.000001770301],"category_scores_gemma":[0.002429214,0.00009899089,0.0001037482,0.0002233127,0.00003742663,0.0000587479,0.0001054995,0.0001069059,0.000004570278],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0002515433,"about_ca_system_score_gemma":0.000172682,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.00006228605,"about_ca_topic_score_gemma":0.000004420762,"domain_scores_codex":[0.9983521,0.00009610302,0.0004525976,0.0002505305,0.0006175733,0.0002310682],"domain_scores_gemma":[0.9957606,0.001423124,0.0006281431,0.0002210978,0.001922886,0.0000441257],"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":[0.000003902109,0.00002843924,0.002832847,0.00002727017,0.00003416657,8.008526e-8,0.0001514688,0.6463704,0.00001064594,0.3503614,0.0001482595,0.00003106172],"study_design_scores_gemma":[0.0004305147,0.0000246347,0.001169655,0.00006612679,0.0000188537,7.859666e-7,0.00001708184,0.7890018,0.00004683448,0.2083533,0.0007629225,0.0001075376],"study_design_candidate":"simulation_or_modeling","study_design_consensus":"simulation_or_modeling","genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.04164557,0.00001607823,0.955977,0.0006329257,0.0004862393,0.0005195091,0.00003231192,0.00006673187,0.0006236414],"genre_scores_gemma":[0.7730668,2.024573e-7,0.2265513,0.00005948116,0.0001437257,0.00003353256,0.00002706426,0.000009068507,0.0001088294],"genre_candidate":"methods","genre_consensus":null,"teacher_disagreement_score":0.7314212,"threshold_uncertainty_score":0.4036732,"prediction_status":"machine_predicted_unvalidated"},"labels":[],"label_agreement":null},{"id":"W2250341893","doi":"","title":"Mapping Source to Target Strings without Alignment by Analogical Learning: A Case Study with Transliteration","year":2013,"lang":"en","type":"article","venue":"Meeting of the Association for Computational Linguistics","topic":"Natural Language Processing Techniques","field":"Computer Science","cited_by":7,"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":"Transliteration; Computer science; Natural language processing; Phrase; Analogy; Artificial intelligence; Machine translation; Task (project management); Translation (biology); Rule-based machine translation; Machine learning; Linguistics; Engineering","authors":[{"name":"Phillippe Langlais","is_ca":true}],"retraction":null,"screen_n_in":null,"score":{"opus":0.01025360225145763,"gpt":0.2590147617676204,"spread":0.2487611595161628,"validation_status":"score_only:v0-immature-baseline"},"prediction":{"model_version":"codex-gemma-dda1882f352a","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.000757726,0.0001320988,0.0001810075,0.0000800094,0.000333396,0.0002263061,0.0003702122,0.00005213444,0.00000100011],"category_scores_gemma":[0.002252002,0.0000987197,0.00005499671,0.000355093,0.00001355505,0.00007125783,0.00009918672,0.0001501247,0.000001229178],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0001762866,"about_ca_system_score_gemma":0.00005026969,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.0001523284,"about_ca_topic_score_gemma":0.000006303101,"domain_scores_codex":[0.9984739,0.0001436029,0.0003660218,0.0002668097,0.0005601733,0.0001895214],"domain_scores_gemma":[0.9976861,0.0004906527,0.0004878408,0.0001356411,0.001147687,0.00005201483],"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":[0.00006214707,0.001071251,0.3736756,0.0001979171,0.0004418256,0.00003789694,0.03541888,0.5595773,0.001705689,0.01803501,0.004299872,0.005476594],"study_design_scores_gemma":[0.002136267,0.00155191,0.003271186,0.0003254478,0.00009856865,0.0001001048,0.002440512,0.9403546,0.003460616,0.03530139,0.01006642,0.0008929443],"study_design_candidate":"simulation_or_modeling","study_design_consensus":"simulation_or_modeling","genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.1377992,0.00003268671,0.8605763,0.0004071346,0.00009781295,0.0008243558,0.000007505376,0.0001568424,0.00009817957],"genre_scores_gemma":[0.6387935,5.523912e-8,0.3608079,0.0001048031,0.00005279933,0.00004187726,0.000009286749,0.000008166116,0.0001816025],"genre_candidate":"methods","genre_consensus":null,"teacher_disagreement_score":0.5009943,"threshold_uncertainty_score":0.4025673,"prediction_status":"machine_predicted_unvalidated"},"labels":[],"label_agreement":null},{"id":"W172656279","doi":"","title":"RALI: Automatic Weighting of Text Window Distances","year":2010,"lang":"en","type":"article","venue":"Meeting of the Association for Computational Linguistics","topic":"Natural Language Processing Techniques","field":"Computer Science","cited_by":5,"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":"Window (computing); Weighting; Word (group theory); Computer science; SemEval; Artificial intelligence; Task (project management); Word-sense disambiguation; Natural language processing; Limit (mathematics); Pattern recognition (psychology); Mathematics","authors":[{"name":"Bernard Brosseau-Villeneuve","is_ca":true},{"name":"Noriko Kando","is_ca":false},{"name":"Jian‐Yun Nie","is_ca":true}],"retraction":null,"screen_n_in":null,"score":{"opus":0.007342829639935821,"gpt":0.2629071066691153,"spread":0.2555642770291795,"validation_status":"score_only:v0-immature-baseline"},"prediction":{"model_version":"codex-gemma-dda1882f352a","candidate_categories":["metaresearch"],"consensus_categories":[],"category_scores_codex":[0.0009596164,0.00009139421,0.0001798814,0.00006291317,0.0001757012,0.00005622371,0.0007422519,0.00007423581,0.000001051821],"category_scores_gemma":[0.01491384,0.000072691,0.0001093479,0.0002876456,0.00004275965,0.00005150101,0.0001450622,0.0001598776,6.145148e-7],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.00005792036,"about_ca_system_score_gemma":0.0001110714,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.0000147355,"about_ca_topic_score_gemma":0.000008912626,"domain_scores_codex":[0.9987249,0.00005272303,0.0004446669,0.0001507553,0.0004846853,0.0001423024],"domain_scores_gemma":[0.9957334,0.001519961,0.001199001,0.0001977843,0.001328847,0.00002095252],"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":[0.000005783078,0.00009038569,0.02613708,0.0002576258,0.00006621148,2.879034e-7,0.0008045856,0.004103833,0.004592216,0.9592136,0.0008483223,0.003880099],"study_design_scores_gemma":[0.0003759511,0.00003908912,0.005801028,0.0002238202,0.0000412475,0.000001304523,0.00001788382,0.3571235,0.0324592,0.6016226,0.002090165,0.000204286],"study_design_candidate":"theoretical_or_conceptual","study_design_consensus":"theoretical_or_conceptual","genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.120207,0.0004358694,0.8637031,0.002014406,0.005930896,0.001008065,0.0001271783,0.0007289007,0.005844598],"genre_scores_gemma":[0.5657179,2.007554e-7,0.4340641,0.00003050363,0.0001106892,0.000003654265,0.000004619036,0.000005103991,0.0000632221],"genre_candidate":"methods","genre_consensus":null,"teacher_disagreement_score":0.4455109,"threshold_uncertainty_score":0.9933839,"prediction_status":"machine_predicted_unvalidated"},"labels":[],"label_agreement":null},{"id":"W1458921027","doi":"","title":"Ontology-Based Extraction and Summarization of Protein Mutation Impact Information","year":2010,"lang":"en","type":"article","venue":"Meeting of the Association for Computational Linguistics","topic":"Biomedical Text Mining and Ontologies","field":"Biochemistry, Genetics and Molecular Biology","cited_by":5,"is_retracted":false,"has_abstract":true,"routes":{"ca_aff":true,"ca_fund":false,"ca_venue":false,"about_ca":false},"ca_institutions":"Concordia University","funders":"","keywords":"Automatic summarization; Computer science; Ontology; Information retrieval; Information extraction; Mutation; Pace; Data mining; Geography; Biology; Genetics","authors":[{"name":"Nona Naderi","is_ca":true},{"name":"René Witte","is_ca":true}],"retraction":null,"screen_n_in":null,"score":{"opus":0.006578853222055952,"gpt":0.277147336279809,"spread":0.2705684830577531,"validation_status":"score_only:v0-immature-baseline"},"prediction":{"model_version":"codex-gemma-dda1882f352a","candidate_categories":["metaresearch"],"consensus_categories":[],"category_scores_codex":[0.0004855969,0.00006245291,0.0000925732,0.00004328782,0.00008252545,0.00001169539,0.00006888044,0.0001488716,8.30727e-7],"category_scores_gemma":[0.01205682,0.00005158483,0.00006383548,0.00007160168,0.00005718931,0.000003250688,0.00001988214,0.00007142694,2.021174e-7],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.00002039361,"about_ca_system_score_gemma":0.000113945,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.00002619178,"about_ca_topic_score_gemma":0.00002112361,"domain_scores_codex":[0.9993622,0.00004088912,0.0002774417,0.00007567706,0.0001676397,0.00007616616],"domain_scores_gemma":[0.9982817,0.0001900044,0.0006281378,0.00007087355,0.000811713,0.00001752522],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"observational","study_design_gemma":"observational","study_design_scores_codex":[0.0008261037,0.0004193689,0.4090079,0.0007220365,0.0004404545,2.146133e-7,0.0006707336,0.1887928,0.3496899,0.02225413,0.002234175,0.0249422],"study_design_scores_gemma":[0.004251902,0.00117134,0.494341,0.0001984986,0.0002271125,0.0000055811,0.0001781844,0.2192615,0.2241382,0.02417159,0.03152517,0.0005299446],"study_design_candidate":"observational","study_design_consensus":"observational","genre_codex":"empirical","genre_gemma":"empirical","genre_scores_codex":[0.9551561,0.00002411616,0.043254,0.0002061119,0.000511244,0.000318917,0.0001185123,0.00001365369,0.0003973019],"genre_scores_gemma":[0.9682922,5.657834e-7,0.03121834,0.00002844188,0.0001101848,0.00000934104,0.0003060826,0.00000452323,0.00003028827],"genre_candidate":"empirical","genre_consensus":"empirical","teacher_disagreement_score":0.1255517,"threshold_uncertainty_score":0.9962651,"prediction_status":"machine_predicted_unvalidated"},"labels":[],"label_agreement":null},{"id":"W3036997889","doi":"","title":"Using Attention-based Bidirectional LSTM to Identify Different Categories of Offensive Language Directed Toward Female Celebrities","year":2019,"lang":"en","type":"article","venue":"Meeting of the Association for Computational Linguistics","topic":"Hate Speech and Cyberbullying Detection","field":"Computer Science","cited_by":4,"is_retracted":false,"has_abstract":true,"routes":{"ca_aff":true,"ca_fund":false,"ca_venue":false,"about_ca":false},"ca_institutions":"Dalhousie University","funders":"","keywords":"Offensive; Harassment; Perspective (graphical); Margin (machine learning); Computer science; Social media; Artificial intelligence; Natural language processing; Psychology; Machine learning; Social psychology; World Wide Web; Mathematics","authors":[{"name":"Sima Sharifirad","is_ca":true},{"name":"Stan Matwin","is_ca":true}],"retraction":null,"screen_n_in":null,"score":{"opus":0.02115456455403126,"gpt":0.2868773790427908,"spread":0.2657228144887596,"validation_status":"score_only:v0-immature-baseline"},"prediction":{"model_version":"codex-gemma-dda1882f352a","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.0004258077,0.0001168365,0.0002267739,0.0001336585,0.0001600161,0.00006305023,0.0002922363,0.00006488995,0.000003361293],"category_scores_gemma":[0.002863056,0.0001075316,0.0001730398,0.0003300072,0.00002044072,0.00002980938,0.00009087651,0.00007995607,0.000005187334],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0002169138,"about_ca_system_score_gemma":0.0001275754,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.0001781424,"about_ca_topic_score_gemma":0.0000139741,"domain_scores_codex":[0.9985248,0.0001065357,0.0004015121,0.0002134293,0.0005818651,0.0001718271],"domain_scores_gemma":[0.9966794,0.0008003659,0.0006069813,0.0001632043,0.00171509,0.00003495542],"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":[0.00009698154,0.0002940639,0.2276473,0.000544872,0.0003424702,9.459251e-7,0.002754455,0.7073027,0.03552566,0.02404655,0.0009538616,0.0004901367],"study_design_scores_gemma":[0.001461315,0.0002356771,0.1983344,0.000478187,0.0001394542,0.000002217569,0.0005803712,0.6941553,0.09419522,0.008846113,0.001054275,0.0005174954],"study_design_candidate":"simulation_or_modeling","study_design_consensus":"simulation_or_modeling","genre_codex":"empirical","genre_gemma":"empirical","genre_scores_codex":[0.9097106,0.00002851866,0.0863158,0.0001652114,0.002650575,0.0004220596,0.0001011529,0.0000979958,0.0005080348],"genre_scores_gemma":[0.9697,2.824295e-7,0.02950771,0.00005785274,0.0001609786,0.000008374888,0.00003858503,0.00001184179,0.0005143996],"genre_candidate":"empirical","genre_consensus":"empirical","teacher_disagreement_score":0.05998931,"threshold_uncertainty_score":0.438501,"prediction_status":"machine_predicted_unvalidated"},"labels":[],"label_agreement":null},{"id":"W3037566932","doi":"","title":"Augmenting Named Entity Recognition with Commonsense Knowledge","year":2019,"lang":"en","type":"article","venue":"Meeting of the Association for Computational Linguistics","topic":"Topic Modeling","field":"Computer Science","cited_by":4,"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; Université du Québec","funders":"","keywords":"Commonsense knowledge; Computer science; Natural language processing; Artificial intelligence; Named-entity recognition; Ambiguity; Knowledge base; Natural language understanding; Natural language; Word embedding; Question answering; Embedding; Task (project management)","authors":[{"name":"Ghaith Dekhili","is_ca":true},{"name":"Ngoc Tan Le","is_ca":true},{"name":"Fatiha Sadat","is_ca":true}],"retraction":null,"screen_n_in":null,"score":{"opus":0.01943582013185866,"gpt":0.2500863071328213,"spread":0.2306504870009626,"validation_status":"score_only:v0-immature-baseline"},"prediction":{"model_version":"codex-gemma-dda1882f352a","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.0009367459,0.00008918781,0.0001490604,0.00005454313,0.000183659,0.00006880239,0.0003482245,0.00005039432,0.00000135193],"category_scores_gemma":[0.002918174,0.00007571669,0.00008362982,0.0002069563,0.00001306232,0.00004373732,0.0001303541,0.0001015164,0.00001312482],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0001693853,"about_ca_system_score_gemma":0.0001208332,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.00002098639,"about_ca_topic_score_gemma":0.00001392382,"domain_scores_codex":[0.9988395,0.0001019977,0.0003221045,0.0002009894,0.0003716649,0.0001637481],"domain_scores_gemma":[0.9965599,0.001249576,0.0006289902,0.0002050933,0.001332169,0.00002423314],"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":[0.00005211794,0.0003186531,0.2976326,0.0003657724,0.0002839759,6.164943e-7,0.002324908,0.5097635,0.0002680079,0.1833749,0.001110772,0.004504156],"study_design_scores_gemma":[0.0008021604,0.00004953228,0.008474004,0.0001818949,0.00003691667,0.000001085016,0.00003481242,0.954672,0.0005363664,0.03263614,0.002412883,0.0001621832],"study_design_candidate":"simulation_or_modeling","study_design_consensus":"simulation_or_modeling","genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.4054849,0.00004651145,0.571449,0.0006029322,0.003372638,0.0008809659,0.00005109243,0.0001794741,0.01793248],"genre_scores_gemma":[0.869102,2.910973e-7,0.1301757,0.0000660162,0.0001492996,0.000006337739,0.00001974482,0.000008584409,0.000472073],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.4636171,"threshold_uncertainty_score":0.3493537,"prediction_status":"machine_predicted_unvalidated"},"labels":[],"label_agreement":null},{"id":"W3088355435","doi":"","title":"PACTE: A colloaborative platform for textual annotation","year":2017,"lang":"en","type":"article","venue":"Meeting of the Association for Computational Linguistics","topic":"Natural Language Processing Techniques","field":"Computer Science","cited_by":1,"is_retracted":false,"has_abstract":false,"routes":{"ca_aff":true,"ca_fund":false,"ca_venue":false,"about_ca":false},"ca_institutions":"Polytechnique Montréal; École de Technologie Supérieure","funders":"","keywords":"Annotation; Computer science; Natural language processing; Artificial intelligence; Information retrieval; World Wide Web","authors":[{"name":"Pierre André Ménard","is_ca":true},{"name":"Caroline Barrière","is_ca":true}],"retraction":null,"screen_n_in":null,"score":{"opus":0.02131663514572495,"gpt":0.3138673364363536,"spread":0.2925507012906286,"validation_status":"score_only:v0-immature-baseline"},"prediction":{"model_version":"codex-gemma-dda1882f352a","candidate_categories":["metaresearch"],"consensus_categories":[],"category_scores_codex":[0.0008937849,0.0001007778,0.0001593866,0.00004963473,0.0009134368,0.0002837513,0.0009097329,0.00008436669,1.944731e-7],"category_scores_gemma":[0.03323898,0.0000841582,0.0001121784,0.0001011039,0.00003839877,0.0001343729,0.0001587277,0.00008584297,5.828159e-7],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0001942158,"about_ca_system_score_gemma":0.0001867531,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.00001553305,"about_ca_topic_score_gemma":0.00001037367,"domain_scores_codex":[0.9989377,0.00002748327,0.0003122544,0.0001861846,0.0003729693,0.0001633614],"domain_scores_gemma":[0.9940622,0.00135009,0.001504112,0.0002669478,0.002793053,0.00002360631],"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":[0.00003243077,0.0000539184,0.003790161,0.00008113184,0.00006872568,1.787504e-7,0.0007238848,0.004790661,0.0001878338,0.9833961,0.004127721,0.002747311],"study_design_scores_gemma":[0.0008541791,0.000105275,0.003609474,0.0001329278,0.00004392448,5.31133e-7,0.00004023646,0.2893769,0.007301095,0.6907694,0.007549315,0.0002167093],"study_design_candidate":"theoretical_or_conceptual","study_design_consensus":"theoretical_or_conceptual","genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.001853143,0.000069834,0.9923406,0.001433612,0.001190281,0.0008191039,0.0002434245,0.0001565662,0.001893452],"genre_scores_gemma":[0.571179,3.365086e-7,0.4282885,0.00006822185,0.0001946906,0.0000257973,0.00002543846,0.000006795501,0.0002112737],"genre_candidate":"methods","genre_consensus":null,"teacher_disagreement_score":0.5693259,"threshold_uncertainty_score":0.9749045,"prediction_status":"machine_predicted_unvalidated"},"labels":[],"label_agreement":null},{"id":"W3197573382","doi":"","title":"Proceedings of the ACL-IJCNLP 2021 Student Research Workshop.","year":2021,"lang":"en","type":"article","venue":"Meeting of the Association for Computational Linguistics","topic":"Dental Research and COVID-19","field":"Dentistry","cited_by":0,"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":"Computer science; Mathematics education; Data science; Psychology","authors":[{"name":"Jad Kabbara","is_ca":true},{"name":"Haitao Lin","is_ca":false},{"name":"Amandalynne Paullada","is_ca":false},{"name":"Jannis Vamvas","is_ca":false}],"retraction":null,"screen_n_in":null,"score":{"opus":0.07961984896349812,"gpt":0.418582325272663,"spread":0.3389624763091649,"validation_status":"score_only:v0-immature-baseline"},"prediction":{"model_version":"codex-gemma-dda1882f352a","candidate_categories":["metaresearch"],"consensus_categories":[],"category_scores_codex":[0.002501147,0.0000886737,0.000182145,0.00005949704,0.0004258574,0.0001136567,0.0005966373,0.0000915498,0.00002540355],"category_scores_gemma":[0.09325187,0.00006768681,0.0002314117,0.0007247775,0.00008763413,0.00002087727,0.000539073,0.0003168211,0.0000121013],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.000434614,"about_ca_system_score_gemma":0.0004590023,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.00002558377,"about_ca_topic_score_gemma":0.0000510734,"domain_scores_codex":[0.9967512,0.0001547894,0.0004697702,0.0002107943,0.002130492,0.0002829086],"domain_scores_gemma":[0.9882107,0.004301444,0.000563797,0.0001679472,0.006698805,0.00005727148],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"observational","study_design_gemma":"observational","study_design_scores_codex":[0.0001581232,0.0008403764,0.5191537,0.0005710551,0.0005545872,0.000004955144,0.001413265,0.02727005,0.002380841,0.3045885,0.1426744,0.0003901788],"study_design_scores_gemma":[0.003697498,0.0001282929,0.49832,0.001078236,0.0002412206,0.000009013801,0.004793567,0.01328665,0.02162461,0.3754712,0.08088764,0.0004620326],"study_design_candidate":"observational","study_design_consensus":"observational","genre_codex":"empirical","genre_gemma":"empirical","genre_scores_codex":[0.9509982,0.0003150201,0.0003985248,0.00233553,0.003591576,0.001041591,0.0003792591,0.00003016648,0.04091017],"genre_scores_gemma":[0.9846882,0.000003914277,0.001334554,0.00008346388,0.0004264939,0.00002008591,0.00002385286,0.00001657036,0.01340289],"genre_candidate":"empirical","genre_consensus":"empirical","teacher_disagreement_score":0.09075072,"threshold_uncertainty_score":0.914386,"prediction_status":"machine_predicted_unvalidated"},"labels":[],"label_agreement":null},{"id":"W3177369674","doi":"","title":"AdElectra - Effective Adversarial Electra","year":2021,"lang":"en","type":"article","venue":"Meeting of the Association for Computational Linguistics","topic":"Classical Philosophy and Thought","field":"Arts and Humanities","cited_by":0,"is_retracted":false,"has_abstract":false,"routes":{"ca_aff":true,"ca_fund":false,"ca_venue":false,"about_ca":false},"ca_institutions":"Université de Montréal","funders":"","keywords":"Adversarial system; Computer science; Artificial intelligence","authors":[{"name":"Makesh Narsimhan Sreedhar","is_ca":false},{"name":"Alessandro Sordoni","is_ca":false},{"name":"Siva Reddy","is_ca":false},{"name":"Aaron Courville","is_ca":true}],"retraction":null,"screen_n_in":null,"score":{"opus":0.01477190681740678,"gpt":0.2364592103238264,"spread":0.2216873035064197,"validation_status":"score_only:v0-immature-baseline"},"prediction":{"model_version":"codex-gemma-dda1882f352a","candidate_categories":["metaresearch"],"consensus_categories":[],"category_scores_codex":[0.0002836641,0.00007630795,0.0001414653,0.00002149781,0.0003498537,0.00005099488,0.00009686113,0.00004303181,0.00001858113],"category_scores_gemma":[0.008463089,0.00006578182,0.0001781814,0.0000620768,0.00004295041,0.00001496233,0.00002850856,0.0001142667,0.000004741783],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0001265187,"about_ca_system_score_gemma":0.00009781642,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.0000097169,"about_ca_topic_score_gemma":0.00003281704,"domain_scores_codex":[0.9991432,0.0000869646,0.0002372763,0.000120508,0.0002793594,0.0001326377],"domain_scores_gemma":[0.9965626,0.001673613,0.0003041136,0.00006742522,0.001373178,0.00001905334],"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":[0.000024897,0.0000737966,0.0003741371,0.00002714288,0.000113231,2.829177e-7,0.000997158,0.00657848,0.00004658992,0.988948,0.002696969,0.0001192576],"study_design_scores_gemma":[0.0007690993,0.00007110136,0.0009003734,0.00005750495,0.0001515442,3.779428e-7,0.00007744307,0.008257603,0.002446247,0.7150604,0.2720577,0.0001506272],"study_design_candidate":"theoretical_or_conceptual","study_design_consensus":"theoretical_or_conceptual","genre_codex":"other","genre_gemma":"empirical","genre_scores_codex":[0.04701041,0.0004169041,0.004654868,0.008312715,0.01993921,0.001688632,0.001337947,0.0002731757,0.9163662],"genre_scores_gemma":[0.9937392,6.066202e-7,0.0009108978,0.0002140056,0.003067908,0.0000102102,0.00007563183,0.00001163406,0.001969872],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.9467288,"threshold_uncertainty_score":0.9998891,"prediction_status":"machine_predicted_unvalidated"},"labels":[],"label_agreement":null}]}