{"meta":{"query_hash":"b4e2a36aa5f9","filters":{"venue":"Proceedings of the 2021 Conference on Empirical Methods in Natural Language Processing"},"cohort_total":19,"direct_labels_cover":0,"predictions_cover":19,"exported":19,"export_cap":100000,"truncated":false,"label_status":"direct model label, unvalidated","prediction_status":"machine_predicted_unvalidated (Codex and Gemma teacher distillation)","score_status":"score_only:v0-immature-baseline","snapshot":{"source":"OpenAlex, pinned release, all 482 partitions","release":"2026-06-24","frame_built":"2026-07-12"},"permalink":"https://metacan.xera.ac/q/b4e2a36aa5f9","api":"https://metacan.xera.ac/api/v1/cohort?venue=Proceedings+of+the+2021+Conference+on+Empirical+Methods+in+Natural+Language+Processing"},"results":[{"id":"W3093808828","doi":"10.18653/v1/2021.emnlp-main.740","title":"BARThez: a Skilled Pretrained French Sequence-to-Sequence Model","year":2021,"lang":"en","type":"preprint","venue":"Proceedings of the 2021 Conference on Empirical Methods in Natural Language Processing","topic":"Natural Language Processing Techniques","field":"Computer Science","cited_by":6,"is_retracted":false,"has_abstract":true,"route_ca_aff":true,"route_ca_fund":false,"route_ca_venue":false,"route_about_ca":false,"ca_institutions":"Polytechnique Montréal","funders":"Centre National de la Recherche Scientifique","keywords":"Automatic summarization; Discriminative model; Computer science; Generative grammar; Benchmark (surveying); Transfer of learning; Artificial intelligence; Natural language processing; Sequence (biology); Code (set theory); Generative model; Field (mathematics); Language model; Machine learning; Programming language; Cartography","score_opus":0.08687262609204593,"score_gpt":0.4277476868274459,"score_spread":0.34087506073539997,"validation_status":"score_only:v0-immature-baseline","prediction":{"id":"W3093808828","genre_codex":"methods","genre_gemma":"methods","domain_codex":null,"domain_gemma":null,"model_version":"metacan-v3-hybrid-931329e0061c","genre_candidate":"methods","genre_consensus":"methods","domain_candidate":null,"domain_consensus":null,"prediction_status":"machine_predicted_unvalidated","genre_scores_codex":[0.09347856,0.0027205902,0.7354111,0.0026084161,0.0012499756,0.00053853326,0.023181658,0.10772115,0.033090025],"genre_scores_gemma":[0.5475124,0.001001092,0.31223044,0.0024439795,0.0004040922,0.0011211239,0.068813704,0.005630249,0.060842942],"study_design_codex":"simulation_or_modeling","study_design_gemma":"bench_or_experimental","domain_scores_codex":[0.9996749,0.00007127652,0.000012151337,0.00013981029,0.00004663186,0.0000552225],"domain_scores_gemma":[0.99946266,0.00027585492,0.000020964382,0.00008421552,0.00012565273,0.000030587067],"candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.0006625992,0.0017134658,0.0006130204,0.00080256414,0.00069071655,0.001179443,0.0024354593,0.001802896,0.013419496],"category_scores_gemma":[0.0022681626,0.0006697215,0.0013425185,0.0006744096,0.0005022141,0.0015283347,0.0009989153,0.0027352923,0.0066099893],"study_design_candidate":"bench_or_experimental","study_design_consensus":null,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_system_candidate":false,"about_ca_system_consensus":false,"study_design_scores_codex":[0.00063182216,0.00026011962,0.0032881165,0.00033551216,0.00028984115,0.00053131924,0.00026649318,0.5248369,0.013813981,0.016055811,0.13977604,0.29991406],"study_design_scores_gemma":[0.000041738884,0.000072207105,0.00046321974,0.000021935062,0.000025824977,0.00009028849,0.00004054637,0.9770685,0.0028330248,0.004732363,0.014582627,0.00002770729],"about_ca_topic_score_codex":0.045783427,"about_ca_topic_score_gemma":0.08969801,"teacher_disagreement_score":0.045783427,"about_ca_system_score_codex":0.0013445319,"about_ca_system_score_gemma":0.0018733605,"threshold_uncertainty_score":0.091033876},"labels":[],"label_agreement":null},{"id":"W3120832022","doi":"10.18653/v1/2021.emnlp-main.526","title":"Towards Zero-Shot Knowledge Distillation for Natural Language Processing","year":2021,"lang":"en","type":"article","venue":"Proceedings of the 2021 Conference on Empirical Methods in Natural Language Processing","topic":"Topic Modeling","field":"Computer Science","cited_by":25,"is_retracted":false,"has_abstract":true,"route_ca_aff":true,"route_ca_fund":false,"route_ca_venue":false,"route_about_ca":false,"ca_institutions":"University of British Columbia","funders":"","keywords":"Computer science; Distillation; Task (project management); Benchmark (surveying); Artificial intelligence; Knowledge transfer; Natural language processing; Transfer of learning; Domain knowledge; Variety (cybernetics); Shot (pellet); Machine learning; Domain (mathematical analysis); Knowledge management; Engineering","score_opus":0.08645918045858457,"score_gpt":0.43361835862493486,"score_spread":0.3471591781663503,"validation_status":"score_only:v0-immature-baseline","prediction":{"id":"W3120832022","genre_codex":"methods","genre_gemma":"empirical","domain_codex":null,"domain_gemma":null,"model_version":"metacan-v3-hybrid-931329e0061c","genre_candidate":"empirical","genre_consensus":null,"domain_candidate":null,"domain_consensus":null,"prediction_status":"machine_predicted_unvalidated","genre_scores_codex":[0.048416253,0.0013433833,0.93659717,0.0013803798,0.00014632053,0.00012622475,0.00059432496,0.006360498,0.0050355666],"genre_scores_gemma":[0.563806,0.0007706717,0.42135248,0.0010590482,0.00021387181,0.0003093253,0.0029601965,0.0008664651,0.008661884],"study_design_codex":"design_other","study_design_gemma":"simulation_or_modeling","domain_scores_codex":[0.9982198,0.0006653931,0.00007379331,0.000416614,0.0004626081,0.00016171952],"domain_scores_gemma":[0.9959189,0.0026801133,0.00014464023,0.00079106604,0.0003164377,0.00014897465],"candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.0028041063,0.0014732534,0.0012835445,0.0010310016,0.00085848494,0.001870214,0.0025915857,0.0023007614,0.003460127],"category_scores_gemma":[0.010266993,0.0005637097,0.00096981105,0.0010608813,0.0022122955,0.0058751446,0.0052913562,0.0050153974,0.0018585117],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_system_candidate":false,"about_ca_system_consensus":false,"study_design_scores_codex":[0.0008926634,0.0005786886,0.0014323447,0.00072885485,0.00016474014,0.00031457175,0.0007015468,0.42160124,0.014236334,0.09028196,0.020902859,0.44816414],"study_design_scores_gemma":[0.00003101061,0.000074786665,0.000090237045,0.000024616886,0.000010643291,0.000048743525,0.00005091702,0.9316118,0.0053439164,0.060348384,0.0023526049,0.0000123802965],"about_ca_topic_score_codex":0.0031935622,"about_ca_topic_score_gemma":0.0052288417,"teacher_disagreement_score":0.003460127,"about_ca_system_score_codex":0.0011542354,"about_ca_system_score_gemma":0.0018328294,"threshold_uncertainty_score":0.014829695},"labels":[],"label_agreement":null},{"id":"W3152698349","doi":"10.18653/v1/2021.emnlp-main.230","title":"Masked Language Modeling and the Distributional Hypothesis: Order Word Matters Pre-training for Little","year":2021,"lang":"en","type":"article","venue":"Proceedings of the 2021 Conference on Empirical Methods in Natural Language Processing","topic":"Topic Modeling","field":"Computer Science","cited_by":182,"is_retracted":false,"has_abstract":true,"route_ca_aff":true,"route_ca_fund":false,"route_ca_venue":false,"route_about_ca":false,"ca_institutions":"McGill University","funders":"","keywords":"Computer science; Word (group theory); Natural language processing; Artificial intelligence; Word order; Language model; Downstream (manufacturing); Order (exchange); Parametric statistics; Linguistics; Mathematics","score_opus":0.07636691541448742,"score_gpt":0.38622034162426394,"score_spread":0.3098534262097765,"validation_status":"score_only:v0-immature-baseline","prediction":{"id":"W3152698349","genre_codex":"methods","genre_gemma":"empirical","domain_codex":null,"domain_gemma":null,"model_version":"metacan-v3-hybrid-931329e0061c","genre_candidate":"empirical","genre_consensus":null,"domain_candidate":null,"domain_consensus":null,"prediction_status":"machine_predicted_unvalidated","genre_scores_codex":[0.32950336,0.0012406091,0.6500712,0.0029224993,0.00050738227,0.00017802021,0.0010625193,0.008094657,0.0064197965],"genre_scores_gemma":[0.83006823,0.00027660967,0.15962629,0.0013476348,0.000116368145,0.00022936844,0.003277213,0.0007923447,0.0042659417],"study_design_codex":"design_other","study_design_gemma":"simulation_or_modeling","domain_scores_codex":[0.9984982,0.00068248133,0.00008020862,0.00051439414,0.00011573195,0.00010901767],"domain_scores_gemma":[0.9919951,0.005831776,0.00019366755,0.0013958872,0.00038590064,0.00019771363],"candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.003336436,0.0013912388,0.00075726624,0.00047893162,0.000642718,0.0013469394,0.0013256845,0.0013878603,0.0043026595],"category_scores_gemma":[0.017700795,0.0006649107,0.00082722306,0.00056051416,0.0009945612,0.0046881726,0.0015085139,0.0043040034,0.0024966174],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_system_candidate":false,"about_ca_system_consensus":false,"study_design_scores_codex":[0.0019108157,0.000808427,0.019880205,0.00063665165,0.00035176516,0.0007194346,0.0013434574,0.17906867,0.09119251,0.020549098,0.02816353,0.6553755],"study_design_scores_gemma":[0.000070769725,0.00032681177,0.0032865894,0.000067104855,0.00006628449,0.00022391166,0.00026301018,0.9231259,0.03162676,0.03601005,0.0048786453,0.000054198354],"about_ca_topic_score_codex":0.0039733904,"about_ca_topic_score_gemma":0.008194104,"teacher_disagreement_score":0.0043026595,"about_ca_system_score_codex":0.00062617305,"about_ca_system_score_gemma":0.0016197698,"threshold_uncertainty_score":0.017644942},"labels":[],"label_agreement":null},{"id":"W3152747197","doi":"10.18653/v1/2021.emnlp-main.234","title":"Linguistic Dependencies and Statistical Dependence","year":2021,"lang":"en","type":"preprint","venue":"Proceedings of the 2021 Conference on Empirical Methods in Natural Language Processing","topic":"Language and cultural evolution","field":"Social Sciences","cited_by":2,"is_retracted":false,"has_abstract":true,"route_ca_aff":true,"route_ca_fund":false,"route_ca_venue":false,"route_about_ca":false,"ca_institutions":"McGill University","funders":"","keywords":"Dependency (UML); Pointwise mutual information; Pointwise; Computer science; Context (archaeology); Natural language processing; ENCODE; Artificial intelligence; Simple (philosophy); Linguistics; Mathematics; Mutual information; Geography","score_opus":0.06719722973920086,"score_gpt":0.44821170864457643,"score_spread":0.38101447890537554,"validation_status":"score_only:v0-immature-baseline","prediction":{"id":"W3152747197","genre_codex":"empirical","genre_gemma":"empirical","domain_codex":null,"domain_gemma":null,"model_version":"metacan-v3-hybrid-931329e0061c","genre_candidate":"empirical","genre_consensus":"empirical","domain_candidate":null,"domain_consensus":null,"prediction_status":"machine_predicted_unvalidated","genre_scores_codex":[0.6530908,0.0026015937,0.3124496,0.002844695,0.00013992,0.000099394674,0.0023632024,0.000889124,0.025521686],"genre_scores_gemma":[0.98194087,0.00037675427,0.014726575,0.00029489578,0.00011006363,0.00008272255,0.0011068683,0.00016732421,0.0011939799],"study_design_codex":"observational","study_design_gemma":"theoretical_or_conceptual","domain_scores_codex":[0.9960517,0.0015487461,0.00024355915,0.0012585376,0.0006652073,0.00023229586],"domain_scores_gemma":[0.94821954,0.040312912,0.0043124254,0.0042595426,0.002130572,0.00076493935],"candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.00415345,0.0004757527,0.0006136427,0.0023329153,0.001136994,0.0016925596,0.000953738,0.0011211748,0.005549612],"category_scores_gemma":[0.0467152,0.0006208778,0.0006850646,0.0023061638,0.0031316024,0.004162072,0.0021715176,0.0027144689,0.0011934646],"study_design_candidate":"theoretical_or_conceptual","study_design_consensus":null,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_system_candidate":false,"about_ca_system_consensus":false,"study_design_scores_codex":[0.001061146,0.0002889983,0.36673334,0.0011687444,0.0011096164,0.0014961825,0.0059725377,0.058479186,0.02531352,0.17417996,0.014106853,0.3500899],"study_design_scores_gemma":[0.000043697888,0.00022245036,0.27397767,0.00024557536,0.00033356322,0.0020324779,0.0013742404,0.1769174,0.006789246,0.5224704,0.015380674,0.00021257922],"about_ca_topic_score_codex":0.0025137903,"about_ca_topic_score_gemma":0.0036582504,"teacher_disagreement_score":0.005549612,"about_ca_system_score_codex":0.0008119974,"about_ca_system_score_gemma":0.0006676134,"threshold_uncertainty_score":0.021965802},"labels":[],"label_agreement":null},{"id":"W3153046263","doi":"10.18653/v1/2021.emnlp-main.168","title":"Neural Path Hunter: Reducing Hallucination in Dialogue Systems via Path Grounding","year":2021,"lang":"en","type":"article","venue":"Proceedings of the 2021 Conference on Empirical Methods in Natural Language Processing","topic":"Topic Modeling","field":"Computer Science","cited_by":75,"is_retracted":false,"has_abstract":true,"route_ca_aff":true,"route_ca_fund":true,"route_ca_venue":false,"route_about_ca":false,"ca_institutions":"Canadian Institute for Advanced Research; University of Alberta","funders":"Alberta Machine Intelligence Institute; Compute Canada; Natural Sciences and Engineering Research Council of Canada; Canadian Institute for Advanced Research","keywords":"Computer science; Path (computing); Security token; Artificial neural network; Focus (optics); Artificial intelligence; Suite; Graph; Deep neural networks; Machine learning; Theoretical computer science; History","score_opus":0.055657809110537304,"score_gpt":0.3775607902916954,"score_spread":0.3219029811811581,"validation_status":"score_only:v0-immature-baseline","prediction":{"id":"W3153046263","genre_codex":"methods","genre_gemma":"methods","domain_codex":null,"domain_gemma":null,"model_version":"metacan-v3-hybrid-931329e0061c","genre_candidate":"methods","genre_consensus":"methods","domain_candidate":null,"domain_consensus":null,"prediction_status":"machine_predicted_unvalidated","genre_scores_codex":[0.29435977,0.003279842,0.6520181,0.0017573764,0.0003331092,0.00063393946,0.0052728457,0.036036152,0.0063089496],"genre_scores_gemma":[0.69971323,0.00041406034,0.2824313,0.00051145384,0.00009473303,0.00033997593,0.0113562895,0.0006843264,0.0044545345],"study_design_codex":"design_other","study_design_gemma":"simulation_or_modeling","domain_scores_codex":[0.99811256,0.0007978775,0.00009269069,0.00057433615,0.00029731196,0.00012518368],"domain_scores_gemma":[0.99218106,0.0052176095,0.00037849028,0.0014478329,0.0005767373,0.00019834601],"candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.0030506905,0.001702592,0.00075171527,0.0014142495,0.0007260411,0.0012895168,0.0025685166,0.0018445388,0.0030455827],"category_scores_gemma":[0.014086668,0.00042429086,0.0008198223,0.0007865705,0.0013593731,0.0043690456,0.0041663293,0.0022975162,0.0012589439],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_system_candidate":false,"about_ca_system_consensus":false,"study_design_scores_codex":[0.0020520363,0.00073881407,0.014669017,0.0017425941,0.0003411685,0.0009658912,0.00316566,0.13105634,0.027203433,0.01077154,0.044622976,0.76267064],"study_design_scores_gemma":[0.00021873467,0.00051346334,0.002225341,0.00007450075,0.000119532066,0.00037526392,0.00094937923,0.9294496,0.019073192,0.03334948,0.013586941,0.00006450953],"about_ca_topic_score_codex":0.0039988896,"about_ca_topic_score_gemma":0.008544529,"teacher_disagreement_score":0.0039988896,"about_ca_system_score_codex":0.00079525117,"about_ca_system_score_gemma":0.001119439,"threshold_uncertainty_score":0.016133785},"labels":[],"label_agreement":null},{"id":"W3197275000","doi":"10.18653/v1/2021.emnlp-main.181","title":"Unsupervised Conversation Disentanglement through Co-Training","year":2021,"lang":"en","type":"article","venue":"Proceedings of the 2021 Conference on Empirical Methods in Natural Language Processing","topic":"Topic Modeling","field":"Computer Science","cited_by":20,"is_retracted":false,"has_abstract":true,"route_ca_aff":true,"route_ca_fund":false,"route_ca_venue":false,"route_about_ca":false,"ca_institutions":"Queen's University","funders":"","keywords":"Conversation; Computer science; Classifier (UML); Session (web analytics); Artificial intelligence; Machine learning; Reinforcement learning; Training set; Speech recognition; World Wide Web","score_opus":0.11382687623688474,"score_gpt":0.4306470469350119,"score_spread":0.3168201706981271,"validation_status":"score_only:v0-immature-baseline","prediction":{"id":"W3197275000","genre_codex":"methods","genre_gemma":"empirical","domain_codex":null,"domain_gemma":null,"model_version":"metacan-v3-hybrid-931329e0061c","genre_candidate":"empirical","genre_consensus":null,"domain_candidate":null,"domain_consensus":null,"prediction_status":"machine_predicted_unvalidated","genre_scores_codex":[0.066704236,0.0010808976,0.92131114,0.00038291048,0.00015653625,0.00024587044,0.0004833981,0.0065861526,0.003048972],"genre_scores_gemma":[0.72213465,0.0003271153,0.26307282,0.0004192645,0.0002217147,0.0005777828,0.0028105723,0.0006790292,0.009757007],"study_design_codex":"design_other","study_design_gemma":"simulation_or_modeling","domain_scores_codex":[0.9972361,0.0009276456,0.00009624664,0.0012210725,0.00028217948,0.00023667744],"domain_scores_gemma":[0.9951775,0.002820145,0.00036513578,0.0006993425,0.0005706384,0.00036717468],"candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.0023654432,0.0023565218,0.0014271357,0.0016353289,0.0008437277,0.0010245763,0.0025154206,0.0015668441,0.003083415],"category_scores_gemma":[0.00864773,0.00075415656,0.0012852412,0.0010280003,0.0008892327,0.0036741951,0.0031955193,0.0040841578,0.002427146],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_system_candidate":false,"about_ca_system_consensus":false,"study_design_scores_codex":[0.0013524259,0.0008854615,0.009288321,0.00040053,0.00027183103,0.00029788818,0.0014338327,0.08755701,0.03732668,0.0038808489,0.010366512,0.84693867],"study_design_scores_gemma":[0.000027792967,0.00011357133,0.0013784858,0.000022860042,0.000039083436,0.00009037356,0.00015422136,0.9798365,0.009430743,0.0056388215,0.0032372598,0.000030272795],"about_ca_topic_score_codex":0.00438437,"about_ca_topic_score_gemma":0.00933098,"teacher_disagreement_score":0.00438437,"about_ca_system_score_codex":0.000758416,"about_ca_system_score_gemma":0.0015311359,"threshold_uncertainty_score":0.012509763},"labels":[],"label_agreement":null},{"id":"W3198457396","doi":"10.18653/v1/2021.emnlp-main.259","title":"Learning from Multiple Noisy Augmented Data Sets for Better Cross-Lingual Spoken Language Understanding","year":2021,"lang":"en","type":"article","venue":"Proceedings of the 2021 Conference on Empirical Methods in Natural Language Processing","topic":"Natural Language Processing Techniques","field":"Computer Science","cited_by":2,"is_retracted":false,"has_abstract":true,"route_ca_aff":true,"route_ca_fund":true,"route_ca_venue":false,"route_about_ca":false,"ca_institutions":"Simon Fraser University","funders":"Natural Sciences and Engineering Research Council of Canada","keywords":"Computer science; Benchmark (surveying); Focus (optics); Noise (video); Code (set theory); Artificial intelligence; Training set; Machine learning; Spoken language; Resource (disambiguation); Noise reduction; Natural language processing; Speech recognition; Image (mathematics); Set (abstract data type)","score_opus":0.14898561212540867,"score_gpt":0.4560963868956432,"score_spread":0.3071107747702345,"validation_status":"score_only:v0-immature-baseline","prediction":{"id":"W3198457396","genre_codex":"methods","genre_gemma":"empirical","domain_codex":null,"domain_gemma":null,"model_version":"metacan-v3-hybrid-931329e0061c","genre_candidate":"empirical","genre_consensus":null,"domain_candidate":null,"domain_consensus":null,"prediction_status":"machine_predicted_unvalidated","genre_scores_codex":[0.061603796,0.0016030018,0.9151431,0.0007920721,0.00040109036,0.00019486608,0.0022261783,0.014572403,0.0034634336],"genre_scores_gemma":[0.35176852,0.00066493405,0.62317735,0.0006669614,0.00019140079,0.0005908122,0.01607111,0.001329834,0.005539086],"study_design_codex":"design_other","study_design_gemma":"simulation_or_modeling","domain_scores_codex":[0.99556446,0.0018552758,0.00026199626,0.0015002948,0.00058381853,0.00023411977],"domain_scores_gemma":[0.991971,0.0037860852,0.00030542005,0.0025725116,0.0011967443,0.00016817024],"candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.0042676004,0.0027015463,0.0020762486,0.0017303984,0.0010560991,0.0031226226,0.0027068355,0.0026442823,0.004371716],"category_scores_gemma":[0.0153661985,0.001053637,0.0024899526,0.0018555411,0.0014065579,0.0058330297,0.006328544,0.005297908,0.005076002],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_system_candidate":false,"about_ca_system_consensus":false,"study_design_scores_codex":[0.00059673615,0.0007786389,0.005311682,0.0007123644,0.0005924959,0.0005149361,0.0017321609,0.10535949,0.04347101,0.0058473935,0.019664919,0.8154182],"study_design_scores_gemma":[0.000066357614,0.00029503735,0.0023280468,0.00013104409,0.00012695693,0.0002544672,0.00097438646,0.9354633,0.025449885,0.019628841,0.015158095,0.00012359682],"about_ca_topic_score_codex":0.005940411,"about_ca_topic_score_gemma":0.010348893,"teacher_disagreement_score":0.005940411,"about_ca_system_score_codex":0.00073327584,"about_ca_system_score_gemma":0.0018832154,"threshold_uncertainty_score":0.022569537},"labels":[],"label_agreement":null},{"id":"W3200130628","doi":"10.18653/v1/2021.emnlp-main.122","title":"Conditional probing: measuring usable information beyond a baseline","year":2021,"lang":"en","type":"article","venue":"Proceedings of the 2021 Conference on Empirical Methods in Natural Language Processing","topic":"Topic Modeling","field":"Computer Science","cited_by":24,"is_retracted":false,"has_abstract":true,"route_ca_aff":false,"route_ca_fund":true,"route_ca_venue":false,"route_about_ca":false,"ca_institutions":"","funders":"Natural Sciences and Engineering Research Council of Canada; National Science Foundation","keywords":"Baseline (sea); USable; Representation (politics); Computer science; Word (group theory); Property (philosophy); Identity (music); Natural language processing; Artificial intelligence; Speech recognition; Mathematics","score_opus":0.05992423051901477,"score_gpt":0.36906375505046807,"score_spread":0.3091395245314533,"validation_status":"score_only:v0-immature-baseline","prediction":{"id":"W3200130628","genre_codex":"methods","genre_gemma":"empirical","domain_codex":null,"domain_gemma":null,"model_version":"metacan-v3-hybrid-931329e0061c","genre_candidate":"empirical","genre_consensus":null,"domain_candidate":null,"domain_consensus":null,"prediction_status":"machine_predicted_unvalidated","genre_scores_codex":[0.34397712,0.0006867602,0.64853764,0.00076061365,0.00005536628,0.0001451458,0.0010068103,0.0014616976,0.0033688403],"genre_scores_gemma":[0.96680796,0.00013689509,0.031435136,0.00019942274,0.000035210716,0.00012453234,0.00075925933,0.00016965819,0.000331786],"study_design_codex":"design_other","study_design_gemma":"simulation_or_modeling","domain_scores_codex":[0.9949567,0.0022296172,0.000307234,0.0013680046,0.00082643016,0.00031208413],"domain_scores_gemma":[0.92483455,0.057534855,0.004006672,0.010535407,0.0019197218,0.0011688009],"candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.008933778,0.0013249208,0.0012642334,0.0013374424,0.00064822217,0.0023125778,0.0016508137,0.002263468,0.0021946274],"category_scores_gemma":[0.06834812,0.0007824252,0.0009894497,0.0014457246,0.00306382,0.010359012,0.0037675377,0.004318867,0.00036004937],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_system_candidate":false,"about_ca_system_consensus":false,"study_design_scores_codex":[0.0061236657,0.0013889676,0.10784878,0.0018218855,0.0016600897,0.00062722474,0.0028148182,0.30286637,0.10793937,0.1411985,0.0052264095,0.32048392],"study_design_scores_gemma":[0.00009316794,0.0013932255,0.02475498,0.00014073051,0.00038847493,0.00039138738,0.0003171033,0.6270015,0.036651038,0.306761,0.0019212407,0.00018602631],"about_ca_topic_score_codex":0.0011137223,"about_ca_topic_score_gemma":0.000939293,"teacher_disagreement_score":0.008933778,"about_ca_system_score_codex":0.001101396,"about_ca_system_score_gemma":0.0007890836,"threshold_uncertainty_score":0.047246933},"labels":[],"label_agreement":null},{"id":"W3200172683","doi":"10.18653/v1/2021.emnlp-main.516","title":"Mind the Context: The Impact of Contextualization in Neural Module Networks for Grounding Visual Referring Expressions","year":2021,"lang":"en","type":"article","venue":"Proceedings of the 2021 Conference on Empirical Methods in Natural Language Processing","topic":"Multimodal Machine Learning Applications","field":"Computer Science","cited_by":3,"is_retracted":false,"has_abstract":true,"route_ca_aff":true,"route_ca_fund":false,"route_ca_venue":false,"route_about_ca":false,"ca_institutions":"McGill University","funders":"","keywords":"Contextualization; Computer science; Generalization; Context (archaeology); Set (abstract data type); Exploit; Parameterized complexity; Artificial intelligence; Test set; Embedding; Cube (algebra); Machine learning; Algorithm; Mathematics; Programming language","score_opus":0.06160955148695172,"score_gpt":0.46076189392771477,"score_spread":0.39915234244076303,"validation_status":"score_only:v0-immature-baseline","prediction":{"id":"W3200172683","genre_codex":"methods","genre_gemma":"empirical","domain_codex":null,"domain_gemma":null,"model_version":"metacan-v3-hybrid-931329e0061c","genre_candidate":"empirical","genre_consensus":null,"domain_candidate":null,"domain_consensus":null,"prediction_status":"machine_predicted_unvalidated","genre_scores_codex":[0.34488717,0.0040735505,0.6210233,0.0019684825,0.00033325836,0.00018598164,0.0009781845,0.012462241,0.014087887],"genre_scores_gemma":[0.882308,0.00056310056,0.11038811,0.0007483577,0.00011304859,0.0001387635,0.001403192,0.00044452705,0.003892906],"study_design_codex":"design_other","study_design_gemma":"simulation_or_modeling","domain_scores_codex":[0.99931526,0.00025700303,0.000025720545,0.00028777297,0.000048246173,0.00006603645],"domain_scores_gemma":[0.99922013,0.000393972,0.00007566831,0.0001666888,0.000098380915,0.000045184384],"candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.0010788332,0.0014869269,0.00057892955,0.00038385327,0.00048575416,0.0011220783,0.0019974483,0.0013547597,0.0034073584],"category_scores_gemma":[0.0044407374,0.0005083755,0.0009924402,0.00047480655,0.00069609267,0.0036504483,0.0016880729,0.002028863,0.00092007074],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_system_candidate":false,"about_ca_system_consensus":false,"study_design_scores_codex":[0.0009555078,0.0003321617,0.008352417,0.00046030956,0.00035089912,0.0005309218,0.0007072735,0.42765597,0.038551174,0.02106779,0.013649513,0.487386],"study_design_scores_gemma":[0.000033627915,0.0001242596,0.0011898692,0.000051446954,0.00010313062,0.00010275706,0.00007210169,0.9757529,0.0063278093,0.011971145,0.0042454205,0.000025547804],"about_ca_topic_score_codex":0.0051611285,"about_ca_topic_score_gemma":0.011362897,"teacher_disagreement_score":0.0051611285,"about_ca_system_score_codex":0.0010249707,"about_ca_system_score_gemma":0.0006630484,"threshold_uncertainty_score":0.011398733},"labels":[],"label_agreement":null},{"id":"W3200285169","doi":"10.18653/v1/2021.emnlp-main.779","title":"PICARD: Parsing Incrementally for Constrained Auto-Regressive Decoding from Language Models","year":2021,"lang":"en","type":"preprint","venue":"Proceedings of the 2021 Conference on Empirical Methods in Natural Language Processing","topic":"Natural Language Processing Techniques","field":"Computer Science","cited_by":12,"is_retracted":false,"has_abstract":true,"route_ca_aff":true,"route_ca_fund":false,"route_ca_venue":false,"route_about_ca":false,"ca_institutions":"McGill University; University of Toronto","funders":"","keywords":"Computer science; Decoding methods; Parsing; Language model; Natural language processing; Artificial intelligence; Code (set theory); Programming language; Compiler; Algorithm; Set (abstract data type)","score_opus":0.07064381815737095,"score_gpt":0.41447000363355646,"score_spread":0.3438261854761855,"validation_status":"score_only:v0-immature-baseline","prediction":{"id":"W3200285169","genre_codex":"methods","genre_gemma":"empirical","domain_codex":null,"domain_gemma":null,"model_version":"metacan-v3-hybrid-931329e0061c","genre_candidate":"empirical","genre_consensus":null,"domain_candidate":null,"domain_consensus":null,"prediction_status":"machine_predicted_unvalidated","genre_scores_codex":[0.007871318,0.00033334832,0.89914805,0.00026699548,0.00014303892,0.00013670104,0.0011475304,0.08873478,0.0022182432],"genre_scores_gemma":[0.14289214,0.0003146261,0.827105,0.0006312913,0.0001262331,0.000502458,0.00692154,0.0158948,0.005611944],"study_design_codex":"design_other","study_design_gemma":"simulation_or_modeling","domain_scores_codex":[0.99823964,0.00059366913,0.000115124356,0.0005737827,0.00034004205,0.00013773405],"domain_scores_gemma":[0.9947523,0.0035426621,0.00014328993,0.0009580164,0.00049850816,0.00010524105],"candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.0019269593,0.0030578387,0.0013845079,0.0012316915,0.0007987178,0.0026646508,0.0031267204,0.0022084592,0.014550982],"category_scores_gemma":[0.015300361,0.0019355732,0.0020474,0.0010493022,0.0012563666,0.003214001,0.003295782,0.004347242,0.008778794],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_system_candidate":false,"about_ca_system_consensus":false,"study_design_scores_codex":[0.00066918816,0.00022550125,0.0023981088,0.00082505273,0.00045979206,0.0006857183,0.000682593,0.24307163,0.028873613,0.027164657,0.07495188,0.6199923],"study_design_scores_gemma":[0.00005583789,0.00004607507,0.00017682115,0.000037222526,0.00004491465,0.00013528508,0.000056083274,0.9584664,0.013253034,0.01973841,0.0079432735,0.000046698016],"about_ca_topic_score_codex":0.009828807,"about_ca_topic_score_gemma":0.026092893,"teacher_disagreement_score":0.014550982,"about_ca_system_score_codex":0.0011609967,"about_ca_system_score_gemma":0.0035152729,"threshold_uncertainty_score":0.04867786},"labels":[],"label_agreement":null},{"id":"W3200853751","doi":"10.18653/v1/2021.emnlp-main.71","title":"Predicting emergent linguistic compositions through time: Syntactic frame extension via multimodal chaining","year":2021,"lang":"en","type":"article","venue":"Proceedings of the 2021 Conference on Empirical Methods in Natural Language Processing","topic":"Language, Metaphor, and Cognition","field":"Psychology","cited_by":6,"is_retracted":false,"has_abstract":true,"route_ca_aff":true,"route_ca_fund":true,"route_ca_venue":false,"route_about_ca":false,"ca_institutions":"Vector Institute; University of Toronto","funders":"Social Sciences and Humanities Research Council of Canada; Natural Sciences and Engineering Research Council of Canada; University of Toronto","keywords":"Computer science; Chaining; Natural language processing; Artificial intelligence; Linguistics; Noun; Psychology","score_opus":0.05792580951332454,"score_gpt":0.4257146922206785,"score_spread":0.36778888270735394,"validation_status":"score_only:v0-immature-baseline","prediction":{"id":"W3200853751","genre_codex":"empirical","genre_gemma":"empirical","domain_codex":null,"domain_gemma":null,"model_version":"metacan-v3-hybrid-931329e0061c","genre_candidate":"empirical","genre_consensus":"empirical","domain_candidate":null,"domain_consensus":null,"prediction_status":"machine_predicted_unvalidated","genre_scores_codex":[0.7503597,0.0008136466,0.24055994,0.0005652127,0.000028325392,0.000058834645,0.0009864772,0.00061388547,0.0060139354],"genre_scores_gemma":[0.96448267,0.00019695172,0.033745248,0.00004380051,0.0000132094465,0.00007063334,0.00077902974,0.000083051025,0.00058536517],"study_design_codex":"design_other","study_design_gemma":"simulation_or_modeling","domain_scores_codex":[0.9995276,0.00020702848,0.000020043617,0.00017436506,0.00003895749,0.000032049316],"domain_scores_gemma":[0.9940592,0.004198589,0.00061237934,0.0005937714,0.0004109825,0.00012512259],"candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.0021436233,0.00043370324,0.0003639641,0.0021070684,0.00052143075,0.0017217842,0.0006400653,0.00074137724,0.00322005],"category_scores_gemma":[0.015719179,0.00048437284,0.0008919575,0.0013159366,0.0011152336,0.006860902,0.0015571816,0.0011738077,0.00060500624],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_system_candidate":false,"about_ca_system_consensus":false,"study_design_scores_codex":[0.00096164434,0.0003683374,0.258062,0.0005854857,0.00048471717,0.0014136186,0.011456941,0.096784376,0.036351524,0.08688787,0.005609671,0.5010338],"study_design_scores_gemma":[0.0000385739,0.00014970897,0.06725992,0.0001136852,0.00013888227,0.0005034218,0.0014568536,0.73042125,0.0076669296,0.18747725,0.004653309,0.00012026184],"about_ca_topic_score_codex":0.0045327866,"about_ca_topic_score_gemma":0.0051165516,"teacher_disagreement_score":0.0045327866,"about_ca_system_score_codex":0.0007737384,"about_ca_system_score_gemma":0.00036232677,"threshold_uncertainty_score":0.011336684},"labels":[],"label_agreement":null},{"id":"W3201341200","doi":"10.18653/v1/2021.emnlp-main.42","title":"Mitigating Language-Dependent Ethnic Bias in BERT","year":2021,"lang":"en","type":"preprint","venue":"Proceedings of the 2021 Conference on Empirical Methods in Natural Language Processing","topic":"Topic Modeling","field":"Computer Science","cited_by":5,"is_retracted":false,"has_abstract":true,"route_ca_aff":true,"route_ca_fund":false,"route_ca_venue":false,"route_about_ca":false,"ca_institutions":"Kootenay Association for Science & Technology","funders":"Ministry of Science and ICT, South Korea; National Research Foundation of Korea; National Research Foundation","keywords":"Categorical variable; Computer science; Ethnic group; German; Turkish; Natural language processing; Metric (unit); Language model; Linguistics; Arabic; Gender bias; Word (group theory); Artificial intelligence; Psychology; Machine learning; Sociology; Social psychology","score_opus":0.13897384373430055,"score_gpt":0.44294323317028,"score_spread":0.30396938943597945,"validation_status":"score_only:v0-immature-baseline","prediction":{"id":"W3201341200","genre_codex":"empirical","genre_gemma":"empirical","domain_codex":null,"domain_gemma":null,"model_version":"metacan-v3-hybrid-931329e0061c","genre_candidate":"empirical","genre_consensus":"empirical","domain_candidate":null,"domain_consensus":null,"prediction_status":"machine_predicted_unvalidated","genre_scores_codex":[0.7023168,0.0005701412,0.2899924,0.00050632353,0.000063073225,0.000077681565,0.00042014185,0.0015619858,0.004491512],"genre_scores_gemma":[0.9588619,0.00009810258,0.0386928,0.00008421125,0.00003480532,0.000037761907,0.0007932257,0.00020682752,0.0011903669],"study_design_codex":"design_other","study_design_gemma":"simulation_or_modeling","domain_scores_codex":[0.99821115,0.0009891189,0.000097777294,0.00026438167,0.00028027841,0.00015735933],"domain_scores_gemma":[0.9901862,0.0063448614,0.0008828929,0.00113197,0.0012325338,0.00022150694],"candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.0042722323,0.0008389969,0.0006941632,0.0010942676,0.0008362728,0.0011044266,0.00074633333,0.00065994536,0.001094152],"category_scores_gemma":[0.016233912,0.00031041572,0.0005035033,0.0011461006,0.000544026,0.0024548722,0.0017495163,0.0010994577,0.0005672492],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_system_candidate":false,"about_ca_system_consensus":false,"study_design_scores_codex":[0.0010598901,0.00031177868,0.18833457,0.00038015627,0.0004619137,0.0006083252,0.0036747633,0.33554143,0.019553743,0.024522182,0.007216659,0.41833454],"study_design_scores_gemma":[0.000028000533,0.000093483875,0.020144563,0.000033033586,0.000061063496,0.00024746856,0.0012205441,0.9478813,0.0074704774,0.018794637,0.0039735045,0.00005197468],"about_ca_topic_score_codex":0.007853147,"about_ca_topic_score_gemma":0.017671999,"teacher_disagreement_score":0.007853147,"about_ca_system_score_codex":0.000648275,"about_ca_system_score_gemma":0.00094556547,"threshold_uncertainty_score":0.022594035},"labels":[],"label_agreement":null},{"id":"W3203338521","doi":"10.18653/v1/2021.emnlp-main.328","title":"Language-Aligned Waypoint (LAW) Supervision for Vision-and-Language Navigation in Continuous Environments","year":2021,"lang":"en","type":"article","venue":"Proceedings of the 2021 Conference on Empirical Methods in Natural Language Processing","topic":"Multimodal Machine Learning Applications","field":"Computer Science","cited_by":39,"is_retracted":false,"has_abstract":true,"route_ca_aff":true,"route_ca_fund":true,"route_ca_venue":false,"route_about_ca":false,"ca_institutions":"Simon Fraser University","funders":"Western Canada Research Grid; Compute Canada; Natural Sciences and Engineering Research Council of Canada; Canadian Institute for Advanced Research","keywords":"Waypoint; Computer science; Task (project management); Human–computer interaction; Embodied cognition; Path (computing); Natural language; Work (physics); Measure (data warehouse); Metric (unit); Shortest path problem; Artificial intelligence; Multimedia; Programming language; Real-time computing; Engineering; Theoretical computer science; Database","score_opus":0.026221236821911142,"score_gpt":0.3976567197951307,"score_spread":0.37143548297321954,"validation_status":"score_only:v0-immature-baseline","prediction":{"id":"W3203338521","genre_codex":"methods","genre_gemma":"empirical","domain_codex":null,"domain_gemma":null,"model_version":"metacan-v3-hybrid-931329e0061c","genre_candidate":"empirical","genre_consensus":null,"domain_candidate":null,"domain_consensus":null,"prediction_status":"machine_predicted_unvalidated","genre_scores_codex":[0.05719918,0.0005740678,0.9377365,0.00016030275,0.00006926459,0.00008649895,0.00018152021,0.0029981532,0.0009946027],"genre_scores_gemma":[0.8040998,0.00024979774,0.1928853,0.00013635113,0.00006686397,0.00015001885,0.0007980163,0.00026179975,0.0013519658],"study_design_codex":"design_other","study_design_gemma":"simulation_or_modeling","domain_scores_codex":[0.9986266,0.00040618813,0.000086078115,0.00043145,0.00033949816,0.00011020735],"domain_scores_gemma":[0.9945878,0.0027768435,0.0006443874,0.0007009289,0.0010105473,0.00027956828],"candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.0018532525,0.00075415586,0.0008261462,0.0006477664,0.0005494103,0.0006954458,0.001807616,0.0010911318,0.0015866841],"category_scores_gemma":[0.013887045,0.00040094234,0.00048577428,0.00055277994,0.0011934999,0.0034524915,0.0019172729,0.0017709639,0.00042993922],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_system_candidate":false,"about_ca_system_consensus":false,"study_design_scores_codex":[0.0010545158,0.00051933114,0.008880778,0.00058774673,0.00013938724,0.00027305106,0.0010587228,0.29672167,0.022810286,0.015727833,0.0071546175,0.6450721],"study_design_scores_gemma":[0.00004525778,0.00030804652,0.0017929305,0.00002693252,0.00002323527,0.00008933361,0.00009969287,0.97357696,0.006775008,0.015802696,0.0014294898,0.000030367803],"about_ca_topic_score_codex":0.010746907,"about_ca_topic_score_gemma":0.014773388,"teacher_disagreement_score":0.010746907,"about_ca_system_score_codex":0.0007616959,"about_ca_system_score_gemma":0.0019461178,"threshold_uncertainty_score":0.021368682},"labels":[],"label_agreement":null},{"id":"W3209039755","doi":"10.18653/v1/2021.emnlp-main.821","title":"IndoNLI: A Natural Language Inference Dataset for Indonesian","year":2021,"lang":"en","type":"article","venue":"Proceedings of the 2021 Conference on Empirical Methods in Natural Language Processing","topic":"Topic Modeling","field":"Computer Science","cited_by":14,"is_retracted":false,"has_abstract":true,"route_ca_aff":false,"route_ca_fund":true,"route_ca_venue":false,"route_about_ca":false,"ca_institutions":"","funders":"Universitas Indonesia; York University; Samsung; National Science Foundation","keywords":"Annotation; Indonesian; Sentence; Test set; Set (abstract data type); Inference; Data set","score_opus":0.06648941278896932,"score_gpt":0.43358885394360996,"score_spread":0.36709944115464066,"validation_status":"score_only:v0-immature-baseline","prediction":{"id":"W3209039755","genre_codex":"dataset","genre_gemma":"empirical","domain_codex":null,"domain_gemma":null,"model_version":"metacan-v3-hybrid-931329e0061c","genre_candidate":"empirical","genre_consensus":null,"domain_candidate":null,"domain_consensus":null,"prediction_status":"machine_predicted_unvalidated","genre_scores_codex":[0.057119317,0.002061597,0.021396972,0.00120106,0.00040400636,0.0008615474,0.87964547,0.014065501,0.023244623],"genre_scores_gemma":[0.038948115,0.00025855814,0.023245199,0.00032630155,0.000050104096,0.0008208487,0.93276006,0.00038241036,0.0032084424],"study_design_codex":"not_applicable","study_design_gemma":"bench_or_experimental","domain_scores_codex":[0.9981651,0.00048574424,0.00027141912,0.0006153373,0.0003463194,0.00011615258],"domain_scores_gemma":[0.9973979,0.0010062184,0.00027621837,0.00061351416,0.00050937984,0.00019681036],"candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.0013626796,0.0014640612,0.00083907286,0.0027366038,0.0012735211,0.0012346931,0.0018473407,0.0016589052,0.010813938],"category_scores_gemma":[0.005044593,0.00048020587,0.0007408656,0.002232438,0.0006360148,0.0021403967,0.00179399,0.002141649,0.011376153],"study_design_candidate":"bench_or_experimental","study_design_consensus":null,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_system_candidate":false,"about_ca_system_consensus":false,"study_design_scores_codex":[0.0007707928,0.00096703053,0.017499242,0.0052464013,0.00024111499,0.0021567543,0.0022569357,0.006330294,0.01786698,0.006519645,0.79559696,0.1445479],"study_design_scores_gemma":[0.00034974242,0.0002511561,0.08519321,0.0006117635,0.00019634474,0.0026465005,0.0026624908,0.050623953,0.022610486,0.009092948,0.8254596,0.0003017908],"about_ca_topic_score_codex":0.013645105,"about_ca_topic_score_gemma":0.032189466,"teacher_disagreement_score":0.013645105,"about_ca_system_score_codex":0.0017446107,"about_ca_system_score_gemma":0.0022054038,"threshold_uncertainty_score":0.036176264},"labels":[],"label_agreement":null},{"id":"W3211384195","doi":"10.18653/v1/2021.emnlp-main.130","title":"Translation-based Supervision for Policy Generation in Simultaneous Neural Machine Translation","year":2021,"lang":"en","type":"article","venue":"Proceedings of the 2021 Conference on Empirical Methods in Natural Language Processing","topic":"Natural Language Processing Techniques","field":"Computer Science","cited_by":6,"is_retracted":false,"has_abstract":true,"route_ca_aff":true,"route_ca_fund":false,"route_ca_venue":false,"route_about_ca":false,"ca_institutions":"Simon Fraser University","funders":"","keywords":"Computer science; Machine translation; Artificial intelligence; Translation (biology); Security token; Oracle; Reinforcement learning; Heuristic; Action (physics); Inference; Machine learning; Quality (philosophy); Sentence; Natural language processing; Programming language","score_opus":0.07715591203049524,"score_gpt":0.42118249622465037,"score_spread":0.3440265841941551,"validation_status":"score_only:v0-immature-baseline","prediction":{"id":"W3211384195","genre_codex":"methods","genre_gemma":"empirical","domain_codex":null,"domain_gemma":null,"model_version":"metacan-v3-hybrid-931329e0061c","genre_candidate":"empirical","genre_consensus":null,"domain_candidate":null,"domain_consensus":null,"prediction_status":"machine_predicted_unvalidated","genre_scores_codex":[0.04783933,0.00020762181,0.94709104,0.00028231554,0.000059693713,0.00010716379,0.00006522608,0.0025220471,0.0018255002],"genre_scores_gemma":[0.80670124,0.00007836626,0.19095181,0.00022003044,0.000049032595,0.00026739034,0.00020674504,0.0002248858,0.001300484],"study_design_codex":"simulation_or_modeling","study_design_gemma":"simulation_or_modeling","domain_scores_codex":[0.99821824,0.00085870316,0.00010274641,0.00043896635,0.0002599096,0.00012151597],"domain_scores_gemma":[0.99009657,0.0068315254,0.00071970036,0.0012139612,0.00089199736,0.0002462267],"candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.0034074537,0.0008492752,0.0010377216,0.00041941274,0.00065733626,0.00066397985,0.0015025609,0.0011441164,0.002449974],"category_scores_gemma":[0.01604421,0.00066008314,0.0004260174,0.0004715534,0.0015695475,0.0020224757,0.0013770038,0.002343032,0.0007340631],"study_design_candidate":"simulation_or_modeling","study_design_consensus":"simulation_or_modeling","about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_system_candidate":false,"about_ca_system_consensus":false,"study_design_scores_codex":[0.0006749986,0.00046356436,0.0021647818,0.00025738354,0.000069693924,0.00019243754,0.00045655272,0.6885932,0.011897061,0.016713846,0.0029887408,0.2755277],"study_design_scores_gemma":[0.000027363596,0.00006393868,0.00010757601,0.0000084568055,0.0000062672125,0.00002236248,0.0000133869835,0.98891544,0.0028190145,0.00761265,0.00039633433,0.000007221902],"about_ca_topic_score_codex":0.002956854,"about_ca_topic_score_gemma":0.004953816,"teacher_disagreement_score":0.0034074537,"about_ca_system_score_codex":0.0010145685,"about_ca_system_score_gemma":0.0022827033,"threshold_uncertainty_score":0.01802051},"labels":[],"label_agreement":null},{"id":"W3211439810","doi":"10.18653/v1/2021.emnlp-main.835","title":"Types of Out-of-Distribution Texts and How to Detect Them","year":2021,"lang":"en","type":"article","venue":"Proceedings of the 2021 Conference on Empirical Methods in Natural Language Processing","topic":"Topic Modeling","field":"Computer Science","cited_by":0,"is_retracted":false,"has_abstract":true,"route_ca_aff":false,"route_ca_fund":true,"route_ca_venue":false,"route_about_ca":false,"ca_institutions":"","funders":"York University","keywords":"Computer science; Categorization; Artificial intelligence; Calibration; Natural language processing; Data mining; Statistics; Mathematics","score_opus":0.0736749441821645,"score_gpt":0.3899331424867225,"score_spread":0.31625819830455804,"validation_status":"score_only:v0-immature-baseline","prediction":{"id":"W3211439810","genre_codex":"empirical","genre_gemma":"empirical","domain_codex":null,"domain_gemma":null,"model_version":"metacan-v3-hybrid-931329e0061c","genre_candidate":"empirical","genre_consensus":"empirical","domain_candidate":null,"domain_consensus":null,"prediction_status":"machine_predicted_unvalidated","genre_scores_codex":[0.6693535,0.008470272,0.26912695,0.0072822515,0.0010589169,0.0008162445,0.015986644,0.007240902,0.02066438],"genre_scores_gemma":[0.82513046,0.0016833829,0.15195207,0.000688033,0.00039259865,0.00038223053,0.014468663,0.0009002696,0.004402319],"study_design_codex":"design_other","study_design_gemma":"simulation_or_modeling","domain_scores_codex":[0.9901248,0.003549391,0.0011089261,0.0022155228,0.0026423715,0.000358915],"domain_scores_gemma":[0.938959,0.04427333,0.005402951,0.0051927674,0.0051868227,0.0009851063],"candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.0057848627,0.0008816297,0.0009364073,0.006984505,0.0010901827,0.0044010617,0.0016023337,0.0022039886,0.0028611177],"category_scores_gemma":[0.07817785,0.00055587,0.0007307596,0.003228913,0.0013248514,0.0074931486,0.002227371,0.0019687517,0.0027951356],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_system_candidate":false,"about_ca_system_consensus":false,"study_design_scores_codex":[0.0010737324,0.00033816983,0.30419278,0.004581575,0.00041786605,0.0025023734,0.008854931,0.007849592,0.025042387,0.023959707,0.06861564,0.5525713],"study_design_scores_gemma":[0.00021220402,0.00035816757,0.24583809,0.0024623813,0.00046283758,0.016059548,0.01956075,0.3659904,0.071207255,0.09008551,0.18728623,0.00047668652],"about_ca_topic_score_codex":0.0014084377,"about_ca_topic_score_gemma":0.0022756108,"teacher_disagreement_score":0.006984505,"about_ca_system_score_codex":0.00075715006,"about_ca_system_score_gemma":0.00072160165,"threshold_uncertainty_score":0.030593693},"labels":[],"label_agreement":null},{"id":"W3213180921","doi":"10.18653/v1/2021.emnlp-main.603","title":"Universal-KD: Attention-based Output-Grounded Intermediate Layer Knowledge Distillation","year":2021,"lang":"en","type":"article","venue":"Proceedings of the 2021 Conference on Empirical Methods in Natural Language Processing","topic":"Topic Modeling","field":"Computer Science","cited_by":18,"is_retracted":false,"has_abstract":true,"route_ca_aff":true,"route_ca_fund":false,"route_ca_venue":false,"route_about_ca":false,"ca_institutions":"University of Waterloo","funders":"","keywords":"Computer science; Layer (electronics); Distillation; Matching (statistics); Interpretability; Projection (relational algebra); Base (topology); Space (punctuation); Architecture; Artificial intelligence; Deep learning; Computer architecture; Machine learning; Algorithm; Mathematics; Nanotechnology; Chromatography; Operating system; Materials science; Chemistry","score_opus":0.07139261601228206,"score_gpt":0.3974343703737435,"score_spread":0.32604175436146143,"validation_status":"score_only:v0-immature-baseline","prediction":{"id":"W3213180921","genre_codex":"methods","genre_gemma":"empirical","domain_codex":null,"domain_gemma":null,"model_version":"metacan-v3-hybrid-931329e0061c","genre_candidate":"empirical","genre_consensus":null,"domain_candidate":null,"domain_consensus":null,"prediction_status":"machine_predicted_unvalidated","genre_scores_codex":[0.017353443,0.0005491757,0.9684529,0.00028498739,0.00015982534,0.00010035972,0.00037450745,0.009335288,0.0033895078],"genre_scores_gemma":[0.5222674,0.00040823076,0.46204996,0.00061678817,0.000105534804,0.00026222182,0.0024816853,0.0007841099,0.011024045],"study_design_codex":"design_other","study_design_gemma":"simulation_or_modeling","domain_scores_codex":[0.99910957,0.00014027383,0.00006651468,0.00033748572,0.00017817947,0.0001680753],"domain_scores_gemma":[0.9987871,0.00041302346,0.00008296138,0.00041826963,0.0002216544,0.00007696138],"candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.0011065565,0.0014869481,0.0012436754,0.0012074496,0.0007569246,0.0015306475,0.0035293717,0.0014901153,0.0072759916],"category_scores_gemma":[0.0040924614,0.00066296116,0.0013472692,0.0014672427,0.0010898384,0.0053373547,0.0043886886,0.0030357013,0.002226051],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_system_candidate":false,"about_ca_system_consensus":false,"study_design_scores_codex":[0.00033267765,0.00023474319,0.0010606756,0.00038991056,0.00013649529,0.00022077038,0.0004677069,0.11240563,0.015390086,0.026427519,0.0129435165,0.8299904],"study_design_scores_gemma":[0.000040734674,0.00008623925,0.00032448946,0.000038821803,0.00006153434,0.00009096831,0.00008379138,0.9381472,0.017062612,0.03694777,0.0070834947,0.000032338427],"about_ca_topic_score_codex":0.007131571,"about_ca_topic_score_gemma":0.010524195,"teacher_disagreement_score":0.0072759916,"about_ca_system_score_codex":0.0011972333,"about_ca_system_score_gemma":0.0020891977,"threshold_uncertainty_score":0.02434063},"labels":[],"label_agreement":null},{"id":"W3214455632","doi":"10.18653/v1/2021.emnlp-main.77","title":"Contextualized Query Embeddings for Conversational Search","year":2021,"lang":"en","type":"article","venue":"Proceedings of the 2021 Conference on Empirical Methods in Natural Language Processing","topic":"Topic Modeling","field":"Computer Science","cited_by":41,"is_retracted":false,"has_abstract":true,"route_ca_aff":false,"route_ca_fund":true,"route_ca_venue":false,"route_about_ca":false,"ca_institutions":"","funders":"Natural Sciences and Engineering Research Council of Canada; Canada First Research Excellence Fund","keywords":"Computer science; Leverage (statistics); Inference; Security token; Relevance (law); Pipeline (software); Information retrieval; Query expansion; Query language; Artificial intelligence","score_opus":0.092333595524727,"score_gpt":0.43953950920379287,"score_spread":0.34720591367906584,"validation_status":"score_only:v0-immature-baseline","prediction":{"id":"W3214455632","genre_codex":"methods","genre_gemma":"empirical","domain_codex":null,"domain_gemma":null,"model_version":"metacan-v3-hybrid-931329e0061c","genre_candidate":"empirical","genre_consensus":null,"domain_candidate":null,"domain_consensus":null,"prediction_status":"machine_predicted_unvalidated","genre_scores_codex":[0.027039591,0.0012971435,0.96456516,0.0005952728,0.00007895151,0.0001178203,0.0012272894,0.0025866535,0.0024922525],"genre_scores_gemma":[0.7218541,0.0009861687,0.26397416,0.00046500043,0.00019568762,0.00043441667,0.0040412676,0.00042863836,0.0076205716],"study_design_codex":"simulation_or_modeling","study_design_gemma":"simulation_or_modeling","domain_scores_codex":[0.9992386,0.00026611466,0.000051209132,0.00023581735,0.00012628762,0.00008193574],"domain_scores_gemma":[0.9986927,0.0006620416,0.000108426066,0.00028351645,0.00019791338,0.000055396045],"candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.0009613278,0.0008798506,0.00079085224,0.00096645224,0.0004169227,0.00109613,0.0016131026,0.0013192791,0.0042360625],"category_scores_gemma":[0.0069629415,0.00048180792,0.0008611562,0.0011284112,0.00068074086,0.004279525,0.0014628662,0.001848169,0.0018623141],"study_design_candidate":"simulation_or_modeling","study_design_consensus":"simulation_or_modeling","about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_system_candidate":false,"about_ca_system_consensus":false,"study_design_scores_codex":[0.00065787585,0.0004651977,0.0035383173,0.00059863273,0.00018583833,0.00031479806,0.0010712699,0.43806502,0.018018061,0.12175751,0.024133425,0.391194],"study_design_scores_gemma":[0.000014445077,0.000053804393,0.00024479596,0.000013639749,0.000020995667,0.00006621602,0.000047213587,0.96408737,0.001342468,0.0311892,0.0029047893,0.000015136694],"about_ca_topic_score_codex":0.008016057,"about_ca_topic_score_gemma":0.010511986,"teacher_disagreement_score":0.008016057,"about_ca_system_score_codex":0.0011765318,"about_ca_system_score_gemma":0.0010077246,"threshold_uncertainty_score":0.015938759},"labels":[],"label_agreement":null},{"id":"W4231122779","doi":"10.18653/v1/2021.emnlp-main.288","title":"Automated Generation of Accurate & Fluent Medical X-ray Reports","year":2021,"lang":"en","type":"article","venue":"Proceedings of the 2021 Conference on Empirical Methods in Natural Language Processing","topic":"Topic Modeling","field":"Computer Science","cited_by":38,"is_retracted":false,"has_abstract":true,"route_ca_aff":true,"route_ca_fund":false,"route_ca_venue":false,"route_about_ca":false,"ca_institutions":"University of Alberta","funders":"","keywords":"Computer science; Fluency; Embedding; Generator (circuit theory); Transformer; Natural language processing; Artificial intelligence; Information retrieval; Linguistics","score_opus":0.09840819144765348,"score_gpt":0.4311545131288209,"score_spread":0.33274632168116747,"validation_status":"score_only:v0-immature-baseline","prediction":{"id":"W4231122779","genre_codex":"methods","genre_gemma":"empirical","domain_codex":null,"domain_gemma":null,"model_version":"metacan-v3-hybrid-931329e0061c","genre_candidate":"empirical","genre_consensus":null,"domain_candidate":null,"domain_consensus":null,"prediction_status":"machine_predicted_unvalidated","genre_scores_codex":[0.045215603,0.0006484048,0.91022784,0.0006249353,0.00023252034,0.00040037834,0.0037484374,0.03680521,0.0020966865],"genre_scores_gemma":[0.28485164,0.00054102193,0.69478774,0.00032869107,0.00020606657,0.00041421544,0.012507989,0.0026614699,0.0037012077],"study_design_codex":"design_other","study_design_gemma":"simulation_or_modeling","domain_scores_codex":[0.99838376,0.00059277157,0.00014191908,0.00039544926,0.0004050018,0.00008110123],"domain_scores_gemma":[0.9920467,0.005440183,0.0005026309,0.0009599765,0.00087776827,0.0001728486],"candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.0021569107,0.0015025832,0.00073806057,0.0014880182,0.0001871821,0.0011448392,0.0017043175,0.0010133171,0.005209121],"category_scores_gemma":[0.01298163,0.00052934623,0.0011580989,0.00062255765,0.00042654687,0.0012692036,0.0016917702,0.0008990154,0.003587427],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_system_candidate":false,"about_ca_system_consensus":false,"study_design_scores_codex":[0.0012124905,0.00037557498,0.006043965,0.0015729514,0.0002024853,0.00293525,0.00089531596,0.09937468,0.063964985,0.009044558,0.04680929,0.7675684],"study_design_scores_gemma":[0.00023567337,0.00031862396,0.0021509847,0.00013216535,0.00014425149,0.0019218021,0.0001877435,0.83553445,0.11786331,0.017765932,0.023665497,0.000079565754],"about_ca_topic_score_codex":0.0006545288,"about_ca_topic_score_gemma":0.0008831666,"teacher_disagreement_score":0.005209121,"about_ca_system_score_codex":0.0004288535,"about_ca_system_score_gemma":0.0008585897,"threshold_uncertainty_score":0.017426252},"labels":[],"label_agreement":null}]}