1
When Gradients Collide: Failure Modes of Multi-Objective Prompt Optimization for LLM Judges
当多目标提示优化遭遇梯度冲突，LLM法官评判会陷入哪些失败模式？这篇论文揭示了关键陷阱
arXiv:2605.26046v1 Announce Type: cross Abstract: Customizing an LLM judge to a specific task or domain often involves optimizing its prompt across mu…