返回全部动态
MemoryForge：为类人LLM智能体合成终身记忆
原标题：MemoryForge: Synthesize Lifelong Memory for Human-Like LLM Agents
AI 摘要
arXiv 论文提出 MemoryForge 框架，用于为 LLM 智能体合成类人的终身记忆。该框架通过记忆条件化替代传统提示方法，包含上下文生成器、生活组织器和多分辨率模拟器三个组件。实验表明，MemoryForge 生成的记忆库能使冻结的 LLM 在角色扮演和用户模拟任务中表现出更类人的行为。
以上摘要由 AI 生成，可能存在误差。事实请以原文为准。
正文节选
Computer Science > Computation and Language Title:MemoryForge: Synthesize Lifelong Memory for Human-Like LLM Agents View PDF HTML (experimental) Abstract:Equipping Large Language Models (LLMs) with human-like personas is crucial for agentic applications, such as role-play and user simulation. Traditional prompt-based methods rely on descriptive conditioning by injecting static textual profiles, which often makes agents show generic behaviors due to a lack of realistic life memory. To
发布时间：2026-08-04 12:00
抓取时间：2026-08-04 12:02
来源机构：arXiv