{"id":"W4404782669","doi":"10.18653/v1/2024.emnlp-main.944","title":"On Efficient Language and Vision Assistants for Visually-Situated Natural Language Understanding: What Matters in Reading and Reasoning","year":2024,"lang":"en","type":"article","venue":"","topic":"Multimodal Machine Learning Applications","field":"Computer Science","cited_by":0,"is_retracted":false,"has_abstract":true,"ca_institutions":"Kootenay Association for Science & Technology","funders":"Korea Advanced Institute of Science and Technology","keywords":"Situated; Computer science; Reading (process); Natural language; Natural (archaeology); Natural language processing; Artificial intelligence; Human–computer interaction; Linguistics","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":[0.003030191,0.0009492402,0.00104536,0.0006187338,0.001155243,0.003550209,0.004230631,0.002109362,0.01162558],"category_scores_gemma":[0.02139528,0.0009741426,0.001624693,0.0008370026,0.001787092,0.01368295,0.003495171,0.00407661,0.005842938],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.00190754,"about_ca_system_score_gemma":0.00394557,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.009944311,"about_ca_topic_score_gemma":0.01955305,"domain_scores_codex":[0.9984459,0.0005745582,0.00008774842,0.0004956371,0.0002280677,0.0001681074],"domain_scores_gemma":[0.9916957,0.004650265,0.0003027293,0.00211963,0.000865289,0.0003664135],"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":[0.0008083606,0.0008913978,0.004291163,0.0007985734,0.0001483806,0.0002395256,0.001062516,0.2343776,0.01071143,0.1181833,0.04654124,0.5819464],"study_design_scores_gemma":[0.00006136251,0.00007792932,0.0004147149,0.0000728693,0.00003735376,0.00009129395,0.0002203562,0.8435432,0.005691866,0.1410705,0.008688658,0.00002995054],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.04065971,0.0007153068,0.938402,0.004372981,0.0001588682,0.0001979556,0.0008389008,0.009649334,0.005004972],"genre_scores_gemma":[0.3023027,0.0007520342,0.6843867,0.0009928223,0.0001318858,0.0003330423,0.002575985,0.001824932,0.006699925],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.01162558,"threshold_uncertainty_score":0.03889149,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.01200450638272115,"score_gpt":0.3344200225874453,"score_spread":0.3224155162047241,"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."}}