GIVE：基于知识图谱启发的真实性外推的大语言模型结构化推理
GIVE: Structured Reasoning of Large Language Models with Knowledge Graph Inspired Veracity Extrapolation
AI总结：
提出GIVE推理方法，融合参数与非参数记忆，通过观察、反思、表达三步引导LLM智能体，无需训练即可在科学等任务上提升准确率并实现完全可解释推理。
摘要：
现有的基于上下文提示或强化学习（RL）来提升大语言模型（LLMs）推理能力的方法，依赖于LLMs的内部知识来生成可靠的思维链（CoT）。然而，无论LLMs的规模大小，某些问题无法在单次前向传播中解决。同时，基于智能体的推理系统需要访问全面的非参数化知识库，这在科学和细分领域中通常成本高昂或不可行。我们提出了Graph Inspired Veracity Extrapolation（GIVE，图谱启发的真实性外推），一种融合参数化和非参数化记忆的新型推理方法，旨在以最少的外部输入提升准确推理。GIVE引导LLM智能体选择最相关的专家数据（观察）、进行针对查询的发散性思考（反思），然后综合这些信息生成最终输出（表达）。大量实验表明了我们框架的以下优势：(1) GIVE提升了各种规模LLMs的性能。(2) 在某些场景下，GIVE使较小的LLMs在科学任务中超越更大、更复杂的模型（GPT3.5T + GIVE > GPT4）。(3) GIVE在科学和开放领域评估中均有效。(4) GIVE是一种免训练方法，使LLMs能够解决超出其训练数据的新问题（准确率提升高达43.5% -> 88.2%）。(5) GIVE允许LLM智能体使用受限（极小）和含噪（极大）的知识源进行推理，支持节点数从135到超过840k的知识图谱（KG）。(6) GIVE的推理过程具有完全可解释性。
英文摘要：
Existing approaches based on context prompting or reinforcement learning (RL) to improve the reasoning capacities of large language models (LLMs) depend on the LLMs' internal knowledge to produce reliable Chain-Of-Thought (CoT). However, no matter the size of LLMs, certain problems cannot be resolved in a single forward pass. Meanwhile, agent-based reasoning systems require access to a comprehensive nonparametric knowledge base, which is often costly or not feasible for use in scientific and niche domains. We present Graph Inspired Veracity Extrapolation (GIVE), a novel reasoning method that merges parametric and non-parametric memories to improve accurate reasoning with minimal external input. GIVE guides the LLM agent to select the most pertinent expert data (observe), engage in query-specific divergent thinking (reflect), and then synthesize this information to produce the final output (speak). Extensive experiments demonstrated the following benefits of our framework: (1) GIVE boosts the performance of LLMs across various sizes. (2) In some scenarios, GIVE allows smaller LLMs to surpass larger, more sophisticated ones in scientific tasks (GPT3.5T + GIVE > GPT4). (3) GIVE is effective on scientific and open-domain assessments. (4) GIVE is a training-free method that enables LLMs to tackle new problems that extend beyond their training data (up to 43.5% -> 88.2%} accuracy improvement). (5) GIVE allows LLM agents to reason using both restricted (very small) and noisy (very large) knowledge sources, accommodating knowledge graphs (KG) ranging from 135 to more than 840k nodes. (6) The reasoning process involved in GIVE is fully interpretable.
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。
