{"id":"W3173783447","doi":"10.18653/v1/2021.acl-long.72","title":"DeCLUTR: Deep Contrastive Learning for Unsupervised Textual Representations","year":2021,"lang":"en","type":"article","venue":"","topic":"Topic Modeling","field":"Computer Science","cited_by":362,"is_retracted":false,"has_abstract":true,"ca_institutions":"University Health Network; Vector Institute; University of Toronto","funders":"National Institutes of Health; Compute Canada; Canadian Institute for Advanced Research","keywords":"Computer science; Natural language processing; Artificial intelligence; Linguistics; Computational linguistics; Volume (thermodynamics); Association (psychology); Joint (building); Deep learning; Cognitive science; Psychology; Philosophy; Epistemology; Engineering","routes":{"ca_aff":true,"ca_fund":true,"ca_venue":false,"about_ca":false,"invisible_to_affiliation_only":false},"retraction":null,"screen":null,"direct_labels":[],"prediction":{"model_version":"metacan-v3-hybrid-931329e0061c","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.001263657,0.00132465,0.0009180878,0.001678834,0.0007188595,0.001844311,0.002768557,0.001331372,0.008955889],"category_scores_gemma":[0.004383939,0.0008025209,0.001160055,0.001590054,0.0006686,0.004901622,0.00247002,0.003273334,0.006484618],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.001008362,"about_ca_system_score_gemma":0.0009345738,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.005396095,"about_ca_topic_score_gemma":0.01242438,"domain_scores_codex":[0.999358,0.0002180189,0.00003455754,0.000222862,0.0001113503,0.00005517026],"domain_scores_gemma":[0.9985093,0.0008536704,0.00006700624,0.0003077256,0.0001947715,0.00006750428],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"design_other","study_design_gemma":"simulation_or_modeling","study_design_scores_codex":[0.0005697056,0.0003328612,0.001206654,0.0005123017,0.0002208141,0.000184618,0.0002644453,0.03556187,0.01261357,0.02635219,0.1382617,0.7839192],"study_design_scores_gemma":[0.0001211037,0.0001115062,0.0005258629,0.00007573316,0.00005584304,0.0001076329,0.00009220035,0.8845556,0.01114882,0.07388984,0.02927219,0.00004367623],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"methods","genre_scores_codex":[0.01498035,0.00281955,0.9251804,0.001094118,0.0005837438,0.0001568,0.006015741,0.04332814,0.005841149],"genre_scores_gemma":[0.2472118,0.001427538,0.6886694,0.001225134,0.0004918542,0.0006046725,0.03292605,0.004430823,0.02301267],"genre_candidate":"methods","genre_consensus":"methods","teacher_disagreement_score":0.008955889,"threshold_uncertainty_score":0.02996045,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.03108129907966992,"score_gpt":0.2870895336430146,"score_spread":0.2560082345633448,"validation_status":"score_only:v0-immature-baseline","note":"Baseline scores from an immature model (maturity gate not passed). Scores rank; they never assert a category."}}