PEG-Tab：表格合成中的采样时间记录修复与发布控制
PEG-Tab: Sampling-Time Record Repair and Release Control for Tabular Synthesis
浏览论文内容
中文总结 AI 辅助
针对预训练表格生成器可能泄露训练记录的问题，提出PEG-Tab后训练框架，通过生成替代候选和校准评分控制发布，显著降低复制风险并保持高效用。
中文摘要 AI 辅助
预训练的表格生成器即使在聚合效用保持较高的情况下，也可能重现训练记录。当重新训练不可用或成本过高时，采样和发布是剩余的干预点。我们提出了PEG-Tab（表格合成的训练后能量引导），一个针对冻结表格生成器的训练后修复和发布控制框架。对于每一生成行，一个生成器原生算子创建两个替代方案。一个共享校准分数比较三个候选，偏好较低风险的记录，并应用最终发布检查。我们为GReaT、CTGAN、TVAE和TabDDPM实例化此接口，而不更新其参数。在五个数据集和四个生成器家族中，PEG-Tab将平均近复制从0.078降至0.027，并将聚合精确复制降至零。相对于3倍的事后过滤器，它在16个迁移设置中的12个中保持更高的效用，并在八个中帕累托支配过滤器。收益集中在复制和邻近相关风险上。
英文摘要
Pretrained tabular generators can reproduce training records even when aggregate utility remains high. When retraining is unavailable or too costly, sampling and release are the remaining intervention points. We present PEG-Tab (Post-Training Energy Guidance for Tabular Synthesis), a post-training repair and release-control framework for frozen tabular generators. For each generated row, a generator-native operator creates two alternatives. A shared calibrated score compares the three candidates, favours lower-risk records, and applies a final release check. We instantiate this interface for GReaT, CTGAN, TVAE, and TabDDPM without updating their parameters. Across five datasets and four generator families, PEG-Tab reduces mean Near Copy from $0.078$ to $0.027$ and lowers aggregate Exact Copy to zero. Relative to a $3\times$ post hoc filter, it retains higher utility in 12 of 16 transfer settings and Pareto-dominates the filter in eight. Gains are concentrated in copy and proximity-related risks.
