Meta 与密歇根州立大学的新论文发现，LLM 构建的推理图质量差异很大，且图中的错误会传导到基于这些图计算出的置信度分数。研究测试了 9 个模型，从给定步骤连线到构建完整图逐步减少辅助，图质量更多与模型家族和图结构相关而非模型规模，Qwen3-32B 在回答子问题上的表现反而不如 Qwen3-8B。作者建议，若基于 LLM 生成的推理图做后续计算，应先检查图本身再信任答案或置信度。
New Meta + Michigan State Univ paper finds that LLM-built reasoning graphs vary widely in quality, and their errors carry into the confidence scores computed from them.
Many methods split a question into a graph of sub-questions, then compute answers or confidence on top. They assume the LLM's graph is correct, which is rarely checked.
They tested 9 models with less and less help, from wiring given steps to building the whole graph. Quality tracked model family and graph shape more than size, and Qwen3-32B answered sub-questions worse than Qwen3-8B.
If you build on LLM-made reasoning graphs, check the graph itself before trusting answers or confidence computed from it.
来源：Rohan Paul · x.com