{"id":"W3132704983","doi":"10.48550/arxiv.2102.12887","title":"Significant Improvements over the State of the Art? A Case Study of the MS MARCO Document Ranking Leaderboard","year":2021,"lang":"en","type":"preprint","venue":"arXiv (Cornell University)","topic":"Topic Modeling","field":"Computer Science","cited_by":5,"is_retracted":false,"has_abstract":true,"ca_institutions":"Microsoft (Canada); University of Waterloo","funders":"","keywords":"Ranking (information retrieval); Metric (unit); Context (archaeology); Computer science; Artificial intelligence; State (computer science); Information retrieval; Order (exchange); sort; Ask price; Machine learning; Learning to rank; Algorithm; Engineering; History; Operations management; Economics","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":["metaresearch"],"consensus_categories":[],"category_scores_codex":[0.01377894,0.0007377259,0.001036347,0.002454605,0.002198359,0.003469903,0.0010685,0.001336832,0.002597817],"category_scores_gemma":[0.04465925,0.0001904037,0.0005645582,0.00364959,0.00168816,0.003802245,0.001483946,0.001998918,0.001276512],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.00159206,"about_ca_system_score_gemma":0.0008763148,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.006970083,"about_ca_topic_score_gemma":0.01381713,"domain_scores_codex":[0.9901423,0.006016559,0.0003669972,0.0008951431,0.002144893,0.0004342037],"domain_scores_gemma":[0.9580694,0.03051013,0.001954667,0.004080821,0.003800578,0.001584289],"domain_codex":null,"domain_gemma":"evaluation","domain_candidate":"evaluation","domain_consensus":null,"study_design_codex":"design_other","study_design_gemma":"observational","study_design_scores_codex":[0.004704906,0.002096359,0.1052333,0.002119992,0.0006998022,0.002775842,0.01078582,0.05270164,0.0144837,0.04573525,0.1376927,0.6209707],"study_design_scores_gemma":[0.001031275,0.006278595,0.2814698,0.0007790406,0.0005302422,0.002895596,0.02049698,0.3463487,0.03431969,0.07493278,0.2302927,0.0006245921],"study_design_candidate":"observational","study_design_consensus":null,"genre_codex":"empirical","genre_gemma":"empirical","genre_scores_codex":[0.9330349,0.006892135,0.01805722,0.008066981,0.0003099001,0.0001908436,0.002386566,0.001654959,0.02940659],"genre_scores_gemma":[0.9665986,0.000691506,0.02499545,0.0004556729,0.0003595804,0.00008517886,0.002043581,0.0003832031,0.004387096],"genre_candidate":"empirical","genre_consensus":"empirical","teacher_disagreement_score":0.9862211,"threshold_uncertainty_score":0.07287085,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.06168247512120693,"score_gpt":0.2001469899934812,"score_spread":0.1384645148722742,"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."}}