GraphTool-Instruction: 通过分解子任务指令革新大语言模型中的图推理
GraphTool-Instruction: Revolutionizing Graph Reasoning in LLMs through Decomposed Subtask Instruction
AI总结：
GraphTool-Instruction通过分解图推理任务为三个子任务，提升LLMs在图推理中的性能，实现SOTA效果。
摘要：
大型语言模型（LLMs）已被证明能够理解基本图属性并解决各种图推理任务。现有方法通过专门设计的任务指令对LLMs进行微调，以理解和执行图推理任务。然而，这些文本指令方法通常表现不佳。受工具学习的启发，研究者提出了工具指令方法，通过特殊工具调用（例如函数、API和模型）解决各种图问题，从而在图推理任务中取得了显著改进。然而，当前的工具指令方法专注于工具信息，忽略了图结构信息，这导致在小规模LLMs（少于13B）上表现显著劣化。为解决这一问题，我们提出了GraphTool-Instruction，一种创新的指令微调方法，将图推理任务分解为三个不同的子任务（即图提取、工具名称识别和工具参数提取），并为每个子任务设计专门的指令。我们的GraphTool-Instruction可以作为插件式提示用于不同的LLMs，而无需微调。此外，基于GraphTool-Instruction，我们开发了一个包含二十个图推理任务的数据集，并基于Llama3-8B创建了一个图推理LLM，称为GraphForge。我们在二十个不同图类型的图推理任务上进行了广泛的实验，并发现GraphTool-Instruction在文本指令和工具指令方法中达到了SOTA。在GTools上微调后，GraphForge相比工具指令增强的GPT-3.5-turbo进一步提高了超过30%，并且与高成本的GPT-4o表现相当。我们的代码和数据可在https://anonymous.4open.science/r/GraphTool-Instruction上获取。
英文摘要：
Large language models (LLMs) have been demonstrated to possess the capabilities to understand fundamental graph properties and address various graph reasoning tasks. Existing methods fine-tune LLMs to understand and execute graph reasoning tasks by specially designed task instructions. However, these Text-Instruction methods generally exhibit poor performance. Inspired by tool learning, researchers propose Tool-Instruction methods to solve various graph problems by special tool calling (e.g., function, API and model), achieving significant improvements in graph reasoning tasks. Nevertheless, current Tool-Instruction approaches focus on the tool information and ignore the graph structure information, which leads to significantly inferior performance on small-scale LLMs (less than 13B). To tackle this issue, we propose GraphTool-Instruction, an innovative Instruction-tuning approach that decomposes the graph reasoning task into three distinct subtasks (i.e., graph extraction, tool name identification and tool parameter extraction), and design specialized instructions for each subtask. Our GraphTool-Instruction can be used as a plug-and-play prompt for different LLMs without fine-tuning. Moreover, building on GraphTool-Instruction, we develop GTools, a dataset that includes twenty graph reasoning tasks, and create a graph reasoning LLM called GraphForge based on Llama3-8B. We conduct extensive experiments on twenty graph reasoning tasks with different graph types (e.g., graph size or graph direction), and we find that GraphTool-Instruction achieves SOTA compared to Text-Instruction and Tool-Instruction methods. Fine-tuned on GTools, GraphForge gets further improvement of over 30% compared to the Tool-Instruction enhanced GPT-3.5-turbo, and it performs comparably to the high-cost GPT-4o. Our codes and data are available at https://anonymous.4open.science/r/GraphTool-Instruction.
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。
