研究探讨 LLM 推理何时能可靠复用中间结果
When Can LLM Reasoning Reliably Reuse Intermediate Results?
DeepSeek 系模型实验数据挺有意思：只给答案就能把第二问成功率从 36% 拉到 57%，做 agent 规划的可以看看。
一项 arXiv 研究分析了 LLM 在多子问题任务中保留中间结果的可靠性条件。在 4 个 DeepSeek 模型上，第二个子问题的平均成功率从无上下文的 36.2% 提升到有第一子问题答案的 57.2%，再加入推理过程则为 58.3%。论文针对本地重试给出正确完成的条件，并界定了平均尝试次数，同时分析检查成本和通过率对总算力的影响。实验覆盖 8 个推理模型，任务为复合数学与算术，代码已在 GitHub 开源。
When Can LLM Reasoning Reliably Reuse Intermediate Results?
Sampling complete solutions gives large language models more chances to solve a problem, but each attempt starts again. For tasks with several subproblems, all required answers must be correct in the same attempt. Retaining intermediate results can avoid repeating correct work, but errors in those results may affect later attempts. We study when this reuse is reliable and how much computation it requires. For local retries that stop when no candidate passes within the retry budget, we give conditions for correct completion and bound the average number of attempts. We also show how checking cost and the chance of passing a check affect total computation. Experiments across eight reasoning models examine composite mathematics and arithmetic. Across four DeepSeek models, average success on the second subproblem is 36.2 percent without first-subproblem context, 57.2 percent with its answer, and 58.3 percent with its reasoning and answer. These scores measure second-subproblem coverage and include continuations made when selection fails. Arithmetic experiments show that similar first-attempt accuracies can lead to very different coverage after repeated attempts. Code is available at https://github.com/SempraETY/Intermediate-State-Reuse.
