arXivDaily arXiv每日学术速递 周一至周五更新
arXiv周末暂无论文更新,休息一下吧,周末愉快~~
arXiv 2403.19962cs.CLcs.AIcs.LG
收藏

通过调优与多分支推理增强低参数大语言模型的通用智能体能力

Enhancing the General Agent Capabilities of Low-Parameter LLMs through Tuning and Multi-Branch Reasoning

Huazhong University of Science and Technology(华中科技大学);Peng Cheng Laboratory(鹏城实验室);DAMO Academy, Alibaba Group(阿里巴巴集团达摩院)

Qinhao Zhou, Zihan Zhang, Xiang Xiang, Ke Wang, Yuchuan Wu, Yongbin Li

AI总结:

本文针对低参数开源LLMs智能体能力不足的问题,提出利用GPT-4构建智能体专用数据进行监督微调,并结合多路径推理与任务分解,在AgentBench五个任务上有效减少了幻觉和格式错误,提升了智能体性能。

摘要:

开源预训练大语言模型(LLMs)展现出强大的语言理解和生成能力,使其在多种任务中取得了显著成功。然而,当它们被用作智能体来处理现实世界中的复杂问题时,其性能远逊于ChatGPT和GPT-4等大型商业模型。作为智能体,LLMs需要具备任务规划、长期记忆以及利用外部工具的能力,才能获得令人满意的表现。已有多种方法被提出用于增强LLMs的智能体能力。一方面,一些方法涉及构建智能体专用数据并对模型进行微调;另一方面,一些方法侧重于设计能够有效激活LLMs推理能力的提示。我们在7B和13B模型上探索了这两种策略。我们提出了一种利用GPT-4构建智能体专用数据的综合方法。通过使用构建的数据进行监督微调,我们发现对于这些参数规模相对较小的模型,监督微调能够显著减少智能体任务中的幻觉输出和格式错误。此外,多路径推理和任务分解等技术可以有效降低问题复杂度,并提升LLMs作为智能体的性能。我们在AgentBench的五个智能体任务上评估了我们的方法,并取得了令人满意的结果。

英文摘要:

Open-source pre-trained Large Language Models (LLMs) exhibit strong language understanding and generation capabilities, making them highly successful in a variety of tasks. However, when used as agents for dealing with complex problems in the real world, their performance is far inferior to large commercial models such as ChatGPT and GPT-4. As intelligent agents, LLMs need to have the capabilities of task planning, long-term memory, and the ability to leverage external tools to achieve satisfactory performance. Various methods have been proposed to enhance the agent capabilities of LLMs. On the one hand, methods involve constructing agent-specific data and fine-tuning the models. On the other hand, some methods focus on designing prompts that effectively activate the reasoning abilities of the LLMs. We explore both strategies on the 7B and 13B models. We propose a comprehensive method for constructing agent-specific data using GPT-4. Through supervised fine-tuning with constructed data, we find that for these models with a relatively small number of parameters, supervised fine-tuning can significantly reduce hallucination outputs and formatting errors in agent tasks. Furthermore, techniques such as multi-path reasoning and task decomposition can effectively decrease problem complexity and enhance the performance of LLMs as agents. We evaluate our method on five agent tasks of AgentBench and achieve satisfactory results.

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。

↑