{"id":"8aa9e9bb-bdc4-4765-8477-f8aa194cda0a","arxiv_id":"1908.00414","paper_version":3,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":1,"one_line_summary":"Semiparametric estimators can suffer from a nonlinear bias and an average nonparametric bias when the first-step estimator is imprecise, and this paper provides two correction methods that work under weaker rate conditions than the standard n^{1/4} requirement.","lead":"This paper studies semiparametric models in which a first-step nonparametric estimate is imprecise, and shows that the resulting estimator can carry two separate non-negligible biases. It proposes a multi-scale jackknife and an analytical bias correction to restore valid inference, with simulations showing improved coverage.","discovery_kind":"extension","skeptic_critique":null,"referee_report":null,"author_rebuttal":null,"desk_editor":null,"rs_alignment":null,"lean_confirmation":null,"pith_extraction":null,"created_at":"2026-08-14T15:58:51.999408+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":null,"supporting_citations":[],"review_version":1}