{"id":"W4287212715","doi":"10.48550/arxiv.2104.06952","title":"The Surprising Performance of Simple Baselines for Misinformation\\n Detection","year":2021,"lang":"","type":"preprint","venue":"arXiv (Cornell University)","topic":"Misinformation and Its Impacts","field":"Social Sciences","cited_by":1,"is_retracted":false,"has_abstract":true,"ca_institutions":"McGill University; Mila - Quebec Artificial Intelligence Institute","funders":"","keywords":"Computer science; Misinformation; Disinformation; Benchmark (surveying); Social media; Machine learning; Baseline (sea); Artificial intelligence; Set (abstract data type); Support vector machine; Language model; Natural language processing; Data mining; Computer security; World Wide Web","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":[0.01103371,0.004626777,0.00237227,0.005726803,0.002315294,0.004151707,0.004153752,0.004782639,0.00385682],"category_scores_gemma":[0.03439595,0.0006854737,0.00128241,0.003563123,0.002236685,0.00981558,0.003143712,0.005113176,0.008595528],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.003075727,"about_ca_system_score_gemma":0.002206116,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.0175311,"about_ca_topic_score_gemma":0.03109888,"domain_scores_codex":[0.9895406,0.003619663,0.0006188678,0.002721704,0.002832622,0.0006665243],"domain_scores_gemma":[0.9738328,0.01279824,0.001516324,0.006900123,0.004191623,0.0007607361],"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":[0.001981311,0.001067124,0.04054678,0.002609397,0.001436473,0.0003351461,0.0006692641,0.03821279,0.01018068,0.0107884,0.1724448,0.7197279],"study_design_scores_gemma":[0.0002172683,0.001048862,0.02416609,0.0007601828,0.000438876,0.001505596,0.001556467,0.8020901,0.04804498,0.04149986,0.07829656,0.000375184],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"empirical","genre_gemma":"empirical","genre_scores_codex":[0.3860281,0.07967604,0.343785,0.02134662,0.01064205,0.001103421,0.03366412,0.0527794,0.07097514],"genre_scores_gemma":[0.7910187,0.005238562,0.1427924,0.003124774,0.001822621,0.0002765553,0.03521406,0.001388046,0.01912436],"genre_candidate":"empirical","genre_consensus":"empirical","teacher_disagreement_score":0.0175311,"threshold_uncertainty_score":0.05835253,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.08755000486076406,"score_gpt":0.2222214377826836,"score_spread":0.1346714329219195,"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."}}