{"id":"W4387076386","doi":"10.48550/arxiv.2309.14181","title":"Q-Bench: A Benchmark for General-Purpose Foundation Models on Low-level Vision","year":2023,"lang":"en","type":"preprint","venue":"arXiv (Cornell University)","topic":"Multimodal Machine Learning Applications","field":"Computer Science","cited_by":20,"is_retracted":false,"has_abstract":true,"ca_institutions":"","funders":"Institute for Catastrophic Loss Reduction","keywords":"Perception; Computer science; Benchmark (surveying); Construct (python library); Correctness; Artificial intelligence; Softmax function; Pipeline (software); Quality (philosophy); Foundation (evidence); Visual perception; Natural language processing; Machine learning; Psychology; Deep learning; Programming language","routes":{"ca_aff":false,"ca_fund":true,"ca_venue":false,"about_ca":false,"invisible_to_affiliation_only":true},"retraction":null,"screen":null,"direct_labels":[],"prediction":{"model_version":"metacan-v3-hybrid-931329e0061c","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.00689728,0.003603692,0.00134812,0.002764477,0.0008735016,0.0035652,0.005766296,0.00430525,0.008180416],"category_scores_gemma":[0.02949342,0.0008117707,0.002636666,0.001918215,0.001224878,0.004557303,0.003720745,0.003323668,0.005442877],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.002774661,"about_ca_system_score_gemma":0.002074304,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.01669319,"about_ca_topic_score_gemma":0.01870085,"domain_scores_codex":[0.995248,0.00153678,0.0004210328,0.0014723,0.0009984673,0.0003233845],"domain_scores_gemma":[0.9904146,0.005431821,0.0004711899,0.001827615,0.001505378,0.0003493812],"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":[0.001813523,0.001117605,0.008505652,0.004832817,0.001305621,0.0005843153,0.0003804973,0.2334348,0.01559376,0.009558611,0.1406136,0.5822592],"study_design_scores_gemma":[0.0002588013,0.0006505848,0.004266514,0.000417566,0.0001453459,0.0004660658,0.0002499567,0.9454266,0.01493909,0.01322386,0.0198462,0.0001093647],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"methods","genre_scores_codex":[0.1826268,0.01840193,0.5645807,0.003091564,0.001570704,0.002694544,0.06731159,0.1272612,0.032461],"genre_scores_gemma":[0.4112552,0.002215702,0.4399238,0.002035478,0.0002500892,0.001731877,0.1314806,0.003675022,0.007432247],"genre_candidate":"methods","genre_consensus":"methods","teacher_disagreement_score":0.01669319,"threshold_uncertainty_score":0.03647679,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.1809078712691692,"score_gpt":0.2562941724986156,"score_spread":0.07538630122944642,"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."}}