Memdoor复现多智能体编程流程
Memdoor对GVS5H方法的复现将编程任务分配给五个使用独立上下文的Qwen3.8-27B角色及一个修复角色，智能体通过文件通信，并由脚本判定测试是否通过。在LiveCodeBench Hard前十道题的五轮测试中，该流程平均得分为9.8/10；原论文团队流程为9.6/10，Fable 5单次调用结果为8.0/10。报告称已公开全部50次运行、判题输出、成本及服务主机信息。
这次复现公开了多智能体编程流程的详细日志；该流程在50次LiveCodeBench Hard尝试中报告了9.8/10的平均分，为通过磁盘文件协作而非共享聊天上下文编排智能体提供了具体参考。
A multi-agent harness for programming problems, written as a Memdoor workflow. Five roles (planner, two ideators, solver, manager) and a repair worker each start from a fresh context and talk only through files on disk. A shell script that runs the samples decides success, not a model. It is the method of GVS5H (Lee, 2026: five Qwen3.8-27B models match Fable 5 on LiveCodeBench Hard).
Rebuilt with the authors' exact prompts on Qwen3.8-27B, their loop scored a mean of 9.8 of 10 over five passes on the first ten LiveCodeBench Hard problems, against 9.6 for their own harness and 8.0 for Fable 5 alone (below, every run published).
The coder wrote these files itself from a description of the method (2026-10-25). Nobody hand-wrote them.
Five passes on the authors' ten (2026-10-31)
GVS5H's own harness, rebuilt as a Memdoor workflow on the same model: Qwen3.8-27B, a 128K reply cap, the authors' prompts and loop. The first ten problems of their LiveCodeBench Hard list, five passes, every program graded on every public and hidden test with their corrected grader.
#Memdoor#GVS5H#Qwen3.8-27B#Fable 5#LiveCodeBench Hard#OpenRouter