{"id":"W4401023825","doi":"10.24963/ijcai.2024/616","title":"StarFT: Robust Fine-tuning of Zero-shot Models via Spuriosity Alignment","year":2024,"lang":"en","type":"article","venue":"","topic":"VLSI and Analog Circuit Testing","field":"Computer Science","cited_by":1,"is_retracted":false,"has_abstract":true,"ca_institutions":"Kootenay Association for Science & Technology","funders":"National Science and Technology Major Project","keywords":"Adaptation (eye); Computer science; Test (biology); Transfer (computing); Parallel computing; Physics","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":[0.0003060719,0.0001228446,0.0001557864,0.00009952794,0.00007658562,0.0001329551,0.0005200905,0.00003955278,0.00003624334],"category_scores_gemma":[0.0000125549,0.0001057461,0.00007972753,0.0003576,0.00002483178,0.0005213172,0.0002365088,0.0001125738,0.00002816104],"about_ca_system_candidate":false,"about_ca_system_consensus":false,"about_ca_system_score_codex":0.00004066391,"about_ca_system_score_gemma":0.00006530511,"about_ca_topic_candidate":false,"about_ca_topic_consensus":false,"about_ca_topic_score_codex":0.0001749618,"about_ca_topic_score_gemma":0.00001803459,"domain_scores_codex":[0.9988186,0.00002769537,0.0002497142,0.0003568824,0.000289877,0.0002571764],"domain_scores_gemma":[0.999373,0.0001016051,0.00003937157,0.0003580793,0.00004589117,0.0000820831],"domain_codex":null,"domain_gemma":null,"domain_candidate":null,"domain_consensus":null,"study_design_codex":"theoretical_or_conceptual","study_design_gemma":"simulation_or_modeling","study_design_scores_codex":[5.085927e-7,0.0001658723,0.0009334381,0.0003092336,0.0001167584,0.0002110819,0.003097891,0.1432539,0.03968613,0.4525495,0.003983611,0.3556921],"study_design_scores_gemma":[0.00006336936,0.00004056913,0.00007766762,0.0001047893,0.000008004891,0.00001775161,0.00001417899,0.9756265,0.00528907,0.01845108,0.0001630653,0.0001439706],"study_design_candidate":"simulation_or_modeling","study_design_consensus":null,"genre_codex":"methods","genre_gemma":"empirical","genre_scores_codex":[0.01026187,0.0003193621,0.971963,0.0003183599,0.0002201718,0.00006701022,0.000001910954,0.0003022287,0.01654612],"genre_scores_gemma":[0.9842958,0.000006406671,0.01491717,0.0001283744,0.00004146981,0.000002822499,0.000001751479,0.000008334909,0.000597826],"genre_candidate":"empirical","genre_consensus":null,"teacher_disagreement_score":0.974034,"threshold_uncertainty_score":0.4312202,"prediction_status":"machine_predicted_unvalidated"},"machine_scores":{"provisional":true,"baseline":true,"maturity_gate_passed":false,"score_opus":0.06075186670880472,"score_gpt":0.2475930412779976,"score_spread":0.1868411745691929,"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."}}