{"id":"W4392903228","doi":"10.1109/icassp48485.2024.10446522","title":"Revisiting the Equivalence of In-Context Learning and Gradient Descent: The Impact of Data Distribution","year":2024,"lang":"en","type":"article","venue":"","topic":"Domain Adaptation and Few-Shot Learning","field":"Computer Science","cited_by":0,"is_retracted":false,"has_abstract":true,"ca_institutions":"University of British Columbia","funders":"","keywords":"Gradient descent; Computer science; Equivalence (formal languages); Softmax function; Covariance; Algorithm; Artificial intelligence; Mathematics; Deep learning; Statistics; Discrete mathematics; Artificial neural network","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":"codex-gemma-dda1882f352a","candidate_categories":[],"consensus_categories":[],"category_scores_codex":[0.001275407,0.00006021841,0.00008643683,0.00003066215,0.00007853672,0.000112948,0.0005431065,0.00001448172,0.00001444148],"category_scores_gemma":[0.0001979546,0.00003034579,0.0000347327,0.0003721832,0.00007706585,0.00037218,0.0003925276,0.0001707612,0.000002542029],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.00002371502,"about_ca_system_score_gemma":0.00004227141,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.000165498,"about_ca_topic_score_gemma":0.000006683073,"domain_scores_codex":[0.9991915,0.0001415622,0.0002056686,0.0001918746,0.0001445994,0.0001247964],"domain_scores_gemma":[0.9991547,0.0004074406,0.00007569925,0.0003097219,0.00003063593,0.00002179344],"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":[0.000006815803,0.0000112689,0.01314549,0.00007413001,0.00002611824,0.000004853398,0.004450068,0.001216519,0.000947414,0.1169119,0.00009302612,0.8631124],"study_design_scores_gemma":[0.0001020332,0.00004359723,0.05244321,0.0002523785,0.000005708104,0.00001631125,0.001154838,0.9437584,0.0001248453,0.0004761307,0.0015601,0.00006243703],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.2728148,0.003846971,0.7202599,0.001440447,0.00008090752,0.0001824476,0.000009672417,0.00005570921,0.001309149],"genre_scores_gemma":[0.9991789,0.0001450745,0.0005694477,0.0000131405,0.00001831184,0.000001115779,0.000006776169,0.000002570259,0.00006469889],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.9425419,"threshold_uncertainty_score":0.1237466,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.05397476545594782,"score_gpt":0.3352681495908358,"score_spread":0.281293384134888,"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."}}