{"id":"W4385565351","doi":"10.18653/v1/2023.acl-long.99","title":"Precise Zero-Shot Dense Retrieval without Relevance Labels","year":2023,"lang":"en","type":"article","venue":"","topic":"Topic Modeling","field":"Computer Science","cited_by":233,"is_retracted":false,"has_abstract":true,"ca_institutions":"University of Waterloo","funders":"Natural Sciences and Engineering Research Council of Canada","keywords":"Relevance (law); Computer science; Similarity (geometry); Relevance feedback; Embedding; Vector space model; Artificial intelligence; Information retrieval; Encoder; Zero (linguistics); Natural language processing; Document retrieval; Vector space; Encoding (memory); Language model; Image retrieval; Pattern recognition (psychology); Mathematics; Image (mathematics)","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":"codex-gemma-dda1882f352a","candidate_categories":["insufficient_payload"],"consensus_categories":[],"category_scores_codex":[0.0004145172,0.000113703,0.0001372666,0.00008923354,0.00008764274,0.0001170633,0.0007765607,0.00005955215,0.00002909887],"category_scores_gemma":[0.0002185251,0.0001007618,0.00004295561,0.0006896489,0.00002209995,0.0003706152,0.0003704748,0.0001280195,0.0008711078],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.00003249569,"about_ca_system_score_gemma":0.00006138702,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.00001561999,"about_ca_topic_score_gemma":0.000007406373,"domain_scores_codex":[0.9985744,0.00004120645,0.0002115304,0.000464299,0.000363006,0.0003456115],"domain_scores_gemma":[0.9987974,0.0001438651,0.00004477411,0.0008381749,0.00007335383,0.000102469],"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":[0.0001609004,0.0002324346,0.01488034,0.0001908275,0.0001312425,0.0006103278,0.007520853,0.01639572,0.08465694,0.4874887,0.1189521,0.2687797],"study_design_scores_gemma":[0.0007338509,0.00008525283,0.001721075,0.00006126642,0.000007652146,0.00003527564,0.00002662712,0.9060167,0.02970572,0.03406801,0.02702721,0.0005113661],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.224589,0.00007105882,0.7636217,0.001468299,0.0006849657,0.0001746638,0.000001204501,0.001249535,0.008139503],"genre_scores_gemma":[0.8676037,0.0000511213,0.1040457,0.0005981622,0.0001023232,0.000007426298,0.000001385071,0.00001659108,0.02757357],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.889621,"threshold_uncertainty_score":0.9999068,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.05173755000251323,"score_gpt":0.2928313881534951,"score_spread":0.2410938381509819,"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."}}