{"id":"W4385573710","doi":"10.18653/v1/2022.emnlp-main.803","title":"SPE: Symmetrical Prompt Enhancement for Fact Probing","year":2022,"lang":"en","type":"article","venue":"","topic":"Topic Modeling","field":"Computer Science","cited_by":5,"is_retracted":false,"has_abstract":true,"ca_institutions":"Mila - Quebec Artificial Intelligence Institute","funders":"","keywords":"Task (project management); Computer science; Object (grammar); Artificial intelligence; Subject (documents); Symmetry (geometry); Task analysis; Natural language processing; Machine learning; Pattern recognition (psychology); Mathematics; Engineering","routes":{"ca_aff":true,"ca_fund":false,"ca_venue":false,"about_ca":false,"invisible_to_affiliation_only":false},"retraction":null,"screen":null,"direct_labels":[],"prediction":{"model_version":"metacan-v3-hybrid-931329e0061c","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.002825541,0.002617667,0.0008410841,0.001294934,0.0005349322,0.001264608,0.001969807,0.001757168,0.008800063],"category_scores_gemma":[0.01484519,0.0005114175,0.001253898,0.001001332,0.000648746,0.005225809,0.00281174,0.003865522,0.006154462],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0005651359,"about_ca_system_score_gemma":0.001544243,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.001721248,"about_ca_topic_score_gemma":0.00356076,"domain_scores_codex":[0.9983051,0.0006331385,0.000119279,0.0006009352,0.0002286612,0.0001129707],"domain_scores_gemma":[0.9922543,0.00464824,0.0003982271,0.001758621,0.0006341321,0.0003064875],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"design_other","study_design_gemma":"bench_or_experimental","study_design_scores_codex":[0.002003278,0.0008519982,0.009913418,0.001062507,0.0002029143,0.0003822522,0.001283729,0.01372596,0.03255378,0.005351597,0.08500611,0.8476624],"study_design_scores_gemma":[0.000813387,0.001941786,0.01414532,0.0002795303,0.0003860943,0.001066152,0.00132424,0.7511744,0.05992177,0.05119793,0.117486,0.0002634844],"study_design_candidate":"bench_or_experimental","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.1490948,0.004253744,0.6668857,0.002292863,0.001214175,0.001273204,0.02138412,0.1437813,0.009820128],"genre_scores_gemma":[0.4946448,0.0009568727,0.4428135,0.001766975,0.0004824935,0.001245203,0.04504959,0.002174052,0.01086649],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.008800063,"threshold_uncertainty_score":0.02943915,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.04301873187770343,"score_gpt":0.2730614404365639,"score_spread":0.2300427085588604,"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."}}