套娃引导器：学习用大语言模型驱动黑盒大语言模型
Matryoshka Pilot: Learning to Drive Black-Box LLMs with LLMs
AI总结：
针对黑盒LLM因不透明性难以提升推理等能力且现有适配方法不可行的问题，提出轻量级白盒LLM控制器Matryoshka Pilot，通过分解任务为中间输出引导黑盒LLM，有效增强其在复杂长期任务中的能力。
摘要：
尽管黑盒大型语言模型（LLMs）具备令人印象深刻的生成能力，但其固有的不透明性阻碍了推理、规划和个性化等能力的进一步提升。现有工作旨在通过特定领域的适配增强LLM能力，这需要对可访问的模型参数进行额外训练，而这对黑盒LLMs来说是不可行的。为应对这一挑战，我们引入Matryoshka Pilot（M-Pilot），一种轻量级白盒LLM控制器，通过将复杂任务分解为一系列中间输出来引导大规模黑盒LLM生成器。具体而言，我们将黑盒LLM视为环境，M-Pilot作为策略通过提示提供中间指导以驱动黑盒LLM。M-Pilot经训练可在迭代交互中调整黑盒LLM的输出使其与偏好对齐，从而实现可控的多轮生成并在优化中间指导方面自我改进。在多种任务上的实证评估表明，我们的方法能有效增强黑盒LLMs在复杂、长期任务中的能力。我们的代码公开于：https://github.com/lichangh20/Matryoshka。
英文摘要：
Despite the impressive generative abilities of black-box large language models (LLMs), their inherent opacity hinders further advancements in capabilities such as reasoning, planning, and personalization. Existing works aim to enhance LLM capabilities via domain-specific adaptation, which require additional training on accessible model parameters, an infeasible option for black-box LLMs. To address this challenge, we introduce Matryoshka Pilot (M-Pilot), a lightweight white-box LLM controller that guides a large-scale black-box LLM generator by decomposing complex tasks into a series of intermediate outputs. Specifically, we consider the black-box LLM as an environment, with M-Pilot serving as a policy to provide intermediate guidance through prompts for driving the black-box LLM. M-Pilot is trained to pivot the outputs of the black-box LLM aligning with preferences during iterative interaction, which enables controllable multi-turn generation and self-improvement in optimizing intermediate guidance. Empirical evaluations on diverse tasks demonstrate that our method effectively enhances the capabilities of black-box LLMs in complex, long-horizon tasks. Our code is publicly available at: https://github.com/lichangh20/Matryoshka.
机构、总结、中文摘要均由AI分析而来，可能存在错误，请以论文正文为准。
