智能DevSecOps技术调研评估 远程兼职

希望调研分析基于大模型（LLM）的自动化安全运维（DevSecOps）技术。具体的，分析，讲解，复现以下论文，代码，评估现有技术的现状，缺陷。论点清晰，论据充分。

漏洞检测：
23, Fuzz4All: Universal Fuzzing with Large Language Models
修复：
24, Domain Knowledge Matters: Improving Prompts with Fix Templates for Repairing Python Type
23, Copiloting the Copilots: Fusing Large Language Models with Completion Engines for Automated Program Repair
23, Impact of Code Language Models on Automated Program Repair
23, RAP-Gen: Retrieval-Augmented Patch Generation with CodeT5 for Automatic Program Repair
23, Pre-trained Model-based Automated Software Vulnerability Repair: How Far are We

交付件：word或ppt形式报告书
评审方式：腾讯会议
预算：暂定5000~10000，协商决定
时间：暂定一周，协商决定