{"id":"W4402683454","doi":"10.18653/v1/2024.repl4nlp-1.21","title":"Whitening Not Recommended for Classification Tasks in LLMs","year":2024,"lang":"en","type":"article","venue":"","topic":"Artificial Intelligence in Law","field":"Social Sciences","cited_by":0,"is_retracted":false,"has_abstract":true,"ca_institutions":"University of Windsor","funders":"","keywords":"Computer science","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":[0.00424521,0.001767109,0.001039324,0.001037281,0.0008250712,0.002253362,0.001276096,0.001985179,0.01600687],"category_scores_gemma":[0.0272369,0.0004164123,0.0009396524,0.0009834039,0.0008749423,0.005302267,0.00242724,0.002380879,0.009357901],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.0006271779,"about_ca_system_score_gemma":0.001034993,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.0012916,"about_ca_topic_score_gemma":0.002451618,"domain_scores_codex":[0.9957293,0.001573157,0.0004670348,0.0009562432,0.0009810705,0.0002931588],"domain_scores_gemma":[0.9858254,0.005612838,0.0007398203,0.005393579,0.002092406,0.0003359608],"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":[0.0023961,0.0007017818,0.004853774,0.001378678,0.0002773115,0.0002578443,0.0005020586,0.02840207,0.06593067,0.01334002,0.06768719,0.8142726],"study_design_scores_gemma":[0.0004331756,0.001710211,0.00862183,0.0003868343,0.0002326261,0.0007732142,0.0006434447,0.6304631,0.2194368,0.05906326,0.07806876,0.0001666679],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.2939208,0.003891967,0.5861494,0.002491954,0.002170322,0.0007558635,0.007909938,0.07181682,0.030893],"genre_scores_gemma":[0.7125993,0.0005150805,0.2534389,0.0009101353,0.0002595962,0.0005806835,0.01505138,0.003999318,0.01264561],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.01600687,"threshold_uncertainty_score":0.05354828,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.2319611445318203,"score_gpt":0.4650844777137296,"score_spread":0.2331233331819093,"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."}}