{"id":"W6929033617","doi":"10.48448/pcmq-qp41","title":"Automatic Evaluation of Generative Models with Instruction Tuning","year":2023,"lang":"en","type":"other","venue":"Open MIND","topic":"Human Motion and Animation","field":"Engineering","cited_by":0,"is_retracted":false,"has_abstract":false,"ca_institutions":"University of British Columbia","funders":"","keywords":"Generative model; Matching (statistics); Feature (linguistics); Generative grammar; Stability (learning theory); Process (computing)","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":[0.001871183,0.001964012,0.00123506,0.001473443,0.0005387869,0.00187104,0.0030353,0.002345507,0.02209495],"category_scores_gemma":[0.01082155,0.0007288238,0.001013784,0.001010666,0.0006850303,0.001999614,0.001971869,0.001482862,0.006622864],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.001145794,"about_ca_system_score_gemma":0.0008737568,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.007333139,"about_ca_topic_score_gemma":0.009793901,"domain_scores_codex":[0.998256,0.0006073746,0.00007511417,0.0004546955,0.0004776471,0.0001290816],"domain_scores_gemma":[0.9947723,0.002986739,0.0001407482,0.001276439,0.0006413223,0.0001825455],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"design_other","study_design_gemma":"not_applicable","study_design_scores_codex":[0.001349773,0.0006432255,0.002537742,0.0003925692,0.0002202432,0.0002146147,0.000122337,0.3015924,0.01562333,0.009519239,0.03742339,0.6303611],"study_design_scores_gemma":[0.00006699394,0.00002855065,0.0001896561,0.000008210354,0.00001384067,0.00002130325,0.00001267613,0.9892485,0.006196944,0.002747343,0.001457746,0.000008207647],"study_design_candidate":"not_applicable","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"methods","genre_scores_codex":[0.0974825,0.0007995687,0.7224681,0.0005034751,0.0004236585,0.0002873665,0.003294367,0.1512781,0.02346284],"genre_scores_gemma":[0.6194927,0.0001862547,0.3505699,0.0002719398,0.00009335397,0.000265958,0.01021122,0.008360348,0.01054832],"genre_candidate":"methods","genre_consensus":"methods","teacher_disagreement_score":0.02209495,"threshold_uncertainty_score":0.07391495,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.09370928327921453,"score_gpt":0.309072205950342,"score_spread":0.2153629226711274,"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."}}