Papers

arXiv 论文:用静态分析提示在生成阶段提升 LLM 代码安全性

Static Analysis Guidance for Secure LLM Code Generation

Why we picked it

训练或搭建代码生成流程的工程师可以看看这篇:不用搭昂贵的迭代修复循环,在生成时把静态分析告警塞进注释或用户消息,DeepSeek-V4.1-flash 在 RiskyPy 上安全率提了约 12 个百分点,token 只花五分之一。还讲清了什么时候这招不灵,省得踩坑。

一篇 arXiv 论文提出在代码生成阶段直接注入静态分析告警的两种策略:Comment(以代码注释形式插入)和 Steering(插入用户消息),当分析器报错时回滚生成并注入发现。相比传统把分析器放进迭代修复循环的做法,生成期引导的 token 开销仅为基线的 0.9–1.9 倍,而迭代方法最高可达 24.1 倍。

Body

实验覆盖三个模型和三个 Python 基准。在 RiskyPy 上,DeepSeek-V4.1-flash 使用 Steering 后功能且安全的程序比例从 52.5% 升到 64.4%,接近最佳迭代方法的 66.2%,但 token 消耗不足其五分之一。逐规则分析显示,局部可修复的规则受益最大,替换 XML 解析器在三个模型中的两个上是净收益。

论文也划出了方法的适用边界:在 SecEvalPlus 上分析器最多只标记 5.6% 的功能正常但不安全的程序,可引导空间有限;当告警与任务需求冲突时模型倾向服从分析器而破坏程序;两个 Qwen2.5 模型在需要重构代码的修复上表现较差。结论是该方法适合搭配能力强的模型、局部修复类规则,并需过滤与任务冲突的告警。

Original · arXiv: DeepSeek

Static Analysis Guidance for Secure LLM Code Generation

Static analyzers are cheap and widely deployed, which makes their reports a candidate feedback signal for secure LLM code generation. Existing approaches place the analyzer inside an iterative repair loop with intermediate steps, which is costly and makes it hard to isolate the report's effect on the model. We introduce two strategies that give the finding to the model during generation. When the analyzer raises an alert, generation is rolled back, and the finding is inserted either as a code comment (Comment) or in the user message (Steering). We compare them with two iterative approaches and a control without the analyzer, across three models and three Python benchmarks, and we measure security using each benchmark's oracle. Guidance during generation is cheap and can be effective, at 0.9 to 1.9 times the baseline's output tokens against up to 24.1 times for iterative approaches. With DeepSeek-V4.1-flash on RiskyPy, Steering raises the share of functional and secure programs from 52.5% to 64.4%, close to the best iterative approach (66.2%) while using less than a fifth of the tokens. A per-rule analysis shows that rules with a local fix benefit most, and replacing the XML parser is net-positive on two of the three models. The study also identifies when guidance should not be applied. Analyzers flag at most 5.6% of the functional but insecure programs on SecEvalPlus, leaving little guidance to act on. When a static analysis alert (warning or error) contradicts the task, models tend to follow the analyzer and break the program. The two Qwen2.5 models often fail when the fix requires restructuring the code. These results indicate that static-analysis guidance is worth applying with a capable model, on rules with a local fix, and after filtering alerts that conflict with the task specification.

Industry Coordinates & Source Verification

This report is categorized under AITOP's "Papers" desk, focusing on China's AI industry and domestic foundation model ecosystem. Within our cross-regional analytical framework, it directly maps to AI developer toolchains, coding agents, and engineering workflows and US-China foundation model benchmarks and commercialization comparison. Key entities and technical topics under continuous tracking include static analysis, DeepSeek, Qwen2.5, code generation.

Primary source verification: published directly by arXiv: DeepSeek with first-party authority.