{"id":"W6891760304","doi":"10.48448/y8tv-6k23","title":"PromptReps: Prompting Large Language Models to Generate Dense and Sparse Representations for Zero-Shot Document Retrieval","year":2024,"lang":"en","type":"other","venue":"Underline Science Inc.","topic":"","field":"","cited_by":0,"is_retracted":false,"has_abstract":true,"ca_institutions":"University of Waterloo","funders":"","keywords":"Document retrieval; Ranking (information retrieval); Construct (python library); Embedding; Language model; Question answering; Security token; Document classification","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":[0.001256658,0.001368985,0.00113169,0.0009070858,0.0004264389,0.001116974,0.002215602,0.001298067,0.004830685],"category_scores_gemma":[0.004882768,0.0005037448,0.0009913103,0.0007803973,0.0007596221,0.003389394,0.002242615,0.00178389,0.00444906],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0007310059,"about_ca_system_score_gemma":0.001290795,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.00426195,"about_ca_topic_score_gemma":0.00866852,"domain_scores_codex":[0.9991394,0.0002797178,0.00005635255,0.0002286477,0.0002114768,0.00008426404],"domain_scores_gemma":[0.9986172,0.000518431,0.000105145,0.0004408673,0.0002318148,0.00008660642],"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":[0.0006041685,0.0004438507,0.001077039,0.000489365,0.0001591202,0.0002902443,0.0003109601,0.07486087,0.04888108,0.01068107,0.03159438,0.8306079],"study_design_scores_gemma":[0.00009168834,0.0002533654,0.0003585792,0.00001949226,0.00003484434,0.0001997393,0.00009043022,0.9573497,0.02353294,0.01112729,0.006884619,0.00005726749],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"methods","genre_scores_codex":[0.02245535,0.0006003596,0.9578596,0.0002599898,0.0001368706,0.0001769948,0.0005396049,0.01639406,0.001577152],"genre_scores_gemma":[0.3165849,0.0005616388,0.6633612,0.0007586518,0.0002367529,0.0005510214,0.004630417,0.001165208,0.01215032],"genre_candidate":"methods","genre_consensus":"methods","teacher_disagreement_score":0.004830685,"threshold_uncertainty_score":0.01616025,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.06875774127856793,"score_gpt":0.3702844002975418,"score_spread":0.3015266590189739,"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."}}