---
title: "博弈论驱动的多智能体框架有效减轻语言模型幻觉"
source: "arXiv AI/CL/LG（论文）"
source_tier: "t2"
url: "https://arxiv.org/abs/2607.08403v1"
permalink: "https://feed.astralor.com/items/itm_06e880b69d2375a43fdd4a56"
published: "2026-07-09T12:28:39.000Z"
score: 81
channel: "ai"
tags: ["G-Frame","OmniChem","化学","博弈论","多智能体"]
---

# 博弈论驱动的多智能体框架有效减轻语言模型幻觉

## AI 点评

> 这篇文章值得所有关注大模型幻觉问题和科学AI应用的研究者阅读。它巧妙地将博弈论与多智能体结合，不仅在化学领域取得显著效果，而且方法论具有通用性，为其他科学领域提供了思路。对于工业界需要部署轻量级模型于专业场景的团队，这个框架尤其有参考价值。

## 速览

轻量级大语言模型在基于规则的科学领域中严重受限于其模仿语言模式而非公理化推理的倾向，导致频繁出现幻觉。本研究提出G-Frame，一种集成贝叶斯和团队博弈原则的自适应多智能体框架，建立了高质量数据合成与模型训练的自动化闭环。通过结构化推理迫使领域约束内化，合成了363,045条思维链和199,589个问答对。由此产生的7B模型OmniChem在自定义基准和ChemBench上达到与GPT-4o mini相当的性能，同时相对于基础架构幻觉减少了79.46%。进一步展示了OmniChem在分子设计和合成规划中的高级能力。这项工作建立了一种利用自适应多智能体克服内在推理缺陷的可扩展范式，为加速科学领域知识发现提供了可行途径。

## 关键点

- 提出G-Frame多智能体框架，结合贝叶斯与团队博弈原理，用于高质量数据合成和模型训练。
- 通过结构化推理内化领域约束，合成了36万+思维链和20万+问答对。
- OmniChem（7B参数）在化学基准测试中与GPT-4o mini性能持平，幻觉减少79.46%。
- 展示了在分子设计和合成规划中的高级应用。

## 全文

The application of lightweight Large Language Models in rule-based scientific domains remains severely limited by their tendency to mimic linguistic patterns rather than reproduce axiomatic reasoning, causing frequent hallucinations. Here, we show that G-Frame, an adaptive multi-agent framework integrating Bayesian and team game principles, establishes an automated closed-loop for high-quality data synthesis and model training. By forcing the internalization of domain constraints through structured reasoning, we synthesized a specialized corpus of 363,045 chains-of-thought and 199,589 question-answer pairs. The resulting 7B model OmniChem achieves performance parity with GPT 4o mini on custom benchmarks and ChemBench while exhibiting a 79.46% reduction in hallucinations relative to its base architecture. We further demonstrate the advanced capabilities of OmniChem in molecular design and synthesis planning. This work establishes a scalable paradigm utilizing adaptive multi-agents to overcome inherent reasoning deficiencies, offering a feasible pathway for accelerating knowledge discovery in specialized scientific fields.

---

引用前请核对原文;本文由星际流动整理:https://feed.astralor.com/items/itm_06e880b69d2375a43fdd4a56