{"id":"W4404782721","doi":"10.18653/v1/2024.emnlp-main.373","title":"Unifying Multimodal Retrieval via Document Screenshot Embedding","year":2024,"lang":"en","type":"article","venue":"","topic":"Topic Modeling","field":"Computer Science","cited_by":16,"is_retracted":false,"has_abstract":true,"ca_institutions":"University of Waterloo","funders":"Natural Sciences and Engineering Research Council of Canada","keywords":"Computer science; Information retrieval; Embedding; Document retrieval; Artificial intelligence; Natural language processing","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":[0.0007328881,0.001529366,0.001074713,0.002345261,0.0003237786,0.001335952,0.001016779,0.0009528711,0.005918308],"category_scores_gemma":[0.004063617,0.0002800213,0.0009206209,0.001512747,0.0006281721,0.004062784,0.002647349,0.0009019356,0.003739689],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0004668874,"about_ca_system_score_gemma":0.0005618602,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.002710026,"about_ca_topic_score_gemma":0.003810714,"domain_scores_codex":[0.9989945,0.0002337404,0.00007555172,0.0003030612,0.0002869756,0.0001061548],"domain_scores_gemma":[0.9984309,0.0005557997,0.00009465473,0.00047729,0.0003681513,0.00007317225],"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":[0.0007863797,0.000352722,0.0008441387,0.0009471594,0.000130315,0.0002290596,0.0004167777,0.01063661,0.09423765,0.004510897,0.03305392,0.8538544],"study_design_scores_gemma":[0.0002604748,0.001422879,0.007266158,0.0002477846,0.0004098133,0.001715981,0.001344142,0.6227817,0.2321372,0.03680307,0.09525526,0.0003554991],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"methods","genre_scores_codex":[0.06496831,0.004792208,0.8893897,0.0007677287,0.0004368491,0.0007619169,0.005290716,0.02246195,0.0111307],"genre_scores_gemma":[0.3800663,0.002952984,0.5707694,0.000867302,0.0004748522,0.0008188623,0.01655329,0.001578662,0.02591841],"genre_candidate":"methods","genre_consensus":"methods","teacher_disagreement_score":0.005918308,"threshold_uncertainty_score":0.0197987,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.02529742333990012,"score_gpt":0.3007426239319106,"score_spread":0.2754452005920105,"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."}}