AI 中文总结
本文提出低秩分解适配器，构建含24个卫星安全案例的防泄露数据集，在Qwen2.5基线对比中验证其性能，为离线卫星安全计划分解提供基准与概念验证。
AI 中文摘要
部分安全开发场景需要将目标映射为有序、可核查计划的局部模型。我们提出了一种低秩分解适配器，并发布了一个案例不相交语料库，包含24个原创分解结果和24个卫星安全案例衍生的83个下一步示例。为防止参考计划泄露，提示包含一个目标和8个打乱的SPARTA候选。评估候选集通过神示包含所有参考技术构建，因此本研究测量的是受控集内的选择而非检索。在6个固定案例和5个解码种子下，不同适配器及提示式Qwen2.5基线的平均分数存在差异；对3个适配器训练种子的单独贪心检查显示出不可忽略的训练方差。在两种紧凑规模下进行的24折留一案例贪心控制同样显示出权衡：15亿参数时，适配器准确率为0.583，而两次提示的准确率为0.480，召回率为0.586，两次提示的召回率为0.660。适配器准确率相对于最强提示基线的案例重采样区间在15亿和7亿参数时跨越零点。适配器在格式合规性上也显著更高，这避免了将所有分数差异归因于选择。自回归下一步准确率在0.5亿、15亿和7亿参数时分别为0.08、0.06和0.29。因此，我们提供了该公式、防泄露数据集、可复现包和描述性分析作为概念验证，而非可靠独立分解系统或通用适配器优势的证据。另有诊断测试用于检测合同复用。
英文摘要
Some security-development settings require local models that map an objective to an ordered, checkable plan. We present a low-rank decomposition adapter and release a case-disjoint corpus with 24 authored decompositions and 83 derived next-step examples across 24 satellite-security cases. To prevent reference-plan leakage, prompts contain one objective and eight shuffled SPARTA candidates. The evaluation candidate sets are constructed with oracle inclusion of every reference technique, so this study measures selection from a controlled set, not retrieval. On six fixed cases and five decoding seeds, mean scores vary across adapters and prompted Qwen2.5 baselines; a separate greedy check over three adapter training seeds shows non-negligible training variance. A 24-fold greedy leave-one-case-out control at the two compact sizes likewise shows tradeoffs: at 1.5B, adapter precision is 0.583 against 0.480 for two-shot prompting, while recall is 0.586 against 0.660. Case-resampling intervals for adapter precision relative to the strongest prompted baseline cross zero at 1.5B and 7B. The adapter is also substantially more format-compliant, which prevents attributing all score differences to selection. Autoregressive next-step accuracy is 0.08, 0.06, and 0.29 at 0.5B, 1.5B, and 7B. We therefore offer the formulation, leakage-controlled dataset, reproducibility package, and descriptive analysis as a proof of concept, not evidence of a reliable stand-alone decomposition system or a general adapter advantage. A separate diagnostic tests contract reuse.
Comments8 pages, 1 figure, 2 tables
