返回探索

Jev 与 Codex Luna 作为代码品味评分器的对比实验

作者运行了一个玩具流水线,将基于文本的评分标准分解为结构化答案,比较 Jev 与 Codex Luna 在 FrontierCode 风格代码品味任务上的评分表现。

ran a toy pipeline comparing jev vs codex luna as graders on FrontierCode style "code taste" tasks: decomposed text-based rubrics into structured answers (e.g. are all changes necessary? do all changes consistently use the abstractions required by the task?) -> run both models in

· 0 次赞在 X 打开