{"id":"W4402671823","doi":"10.18653/v1/2024.acl-long.208","title":"DataDreamer: A Tool for Synthetic Data Generation and Reproducible LLM Workflows","year":2024,"lang":"en","type":"article","venue":"","topic":"Scientific Computing and Data Management","field":"Decision Sciences","cited_by":19,"is_retracted":false,"has_abstract":true,"ca_institutions":"Vector Institute","funders":"Intelligence Advanced Research Projects Activity; Office of the Director of National Intelligence","keywords":"Workflow; Computer science; Synthetic data; Artificial intelligence; Database","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":[0.01222154,0.001824717,0.001137828,0.00273514,0.001703014,0.004238171,0.005275356,0.00138787,0.02810939],"category_scores_gemma":[0.04187668,0.001869752,0.003000662,0.002570292,0.001870092,0.005085477,0.007429315,0.004601233,0.01454622],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.001466534,"about_ca_system_score_gemma":0.005670669,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.004950581,"about_ca_topic_score_gemma":0.006642116,"domain_scores_codex":[0.9950333,0.001423424,0.0006798242,0.0010711,0.001502734,0.0002896964],"domain_scores_gemma":[0.9834802,0.008436594,0.0008368465,0.004607878,0.001880984,0.000757473],"domain_codex":null,"domain_gemma":"reproducibility","domain_candidate":"reproducibility","domain_consensus":null,"study_design_codex":"not_applicable","study_design_gemma":"not_applicable","study_design_scores_codex":[0.001474294,0.0006615592,0.01107419,0.002933183,0.0007210473,0.001006466,0.003457056,0.04538935,0.01553909,0.05243455,0.6386945,0.2266147],"study_design_scores_gemma":[0.0008145571,0.0001952642,0.003374911,0.0004882018,0.0001138037,0.0004206398,0.0005186577,0.328169,0.04029142,0.1016688,0.5234703,0.0004745195],"study_design_candidate":"not_applicable","study_design_consensus":"not_applicable","genre_codex":"methods","genre_gemma":"software","genre_scores_codex":[0.004678942,0.0001487775,0.5029517,0.0009594798,0.0004685249,0.0006224368,0.03325121,0.4513081,0.005610866],"genre_scores_gemma":[0.04840383,0.0003311259,0.7284234,0.001046266,0.0001180313,0.003280065,0.09245051,0.1199981,0.005948648],"genre_candidate":"software","genre_consensus":null,"teacher_disagreement_score":0.9877785,"threshold_uncertainty_score":0.09403533,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.4313535412028034,"score_gpt":0.457528050023825,"score_spread":0.02617450882102157,"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."}}