解决大模型实时部署中的幻觉与过时问题，实现企业级可靠运行。
Unleashing the Potential of Large Language Models: A Blueprint for Real-Time, Enterprise-Ready Deployments
- 构建统一运维架构，融合数据流、持续学习与人工反馈闭环。
- 在医疗金融场景中降低延迟成本，准确率提升且支持审计回滚。
- 提出自适应检索与稀疏路由学习机制，提升模型长期可靠性。
在实时、受监管的环境中部署大语言模型面临知识过时、灾难性遗忘、幻觉及反馈弱等问题。本文提出一种统一的模式驱动型LLMOps架构，集成实时数据摄入、持续学习、检索增强生成（RAG）与人机协同反馈，形成一体化运营流程。四个核心贡献对应成熟软件设计模式：自适应摄入模式编排器（AIPO），在FreshStreamBench上验证；STAR+FAR持续学习方法，采用稀疏时间适配器路由与新鲜度感知重放；SAGE——一种面向服务等级目标（SLO）的自适应检索策略，可预测每查询的段落预算以满足尾部延迟要求；以及基于强化学习的人工反馈驱动收敛阶段。该方案有效缓解延迟-成本-准确率权衡，支持高风险领域如医疗与金融的可审计性和回滚能力。
原文摘要 · Abstract (English)
Large language models deployed in real-time, regulated settings face knowledge staleness, catastrophic forgetting, hallucination, and weak feedback loops. We present a unified, pattern-driven LLMOps architecture integrating real-time data ingestion, continual learning, retrieval-augmented generation (RAG), and human-in-the-loop feedback into a single operational pipeline. Four contributions map to established software design patterns: an adaptive ingestion pattern orchestrator (AIPO) evaluated with FreshStreamBench; STAR+FAR continual learning with sparse temporal adapter routing and freshness-aware replay; SAGE, an SLO-aware adaptive retrieval policy predicting a per-query passage budget to meet tail-latency targets; and an automated feedback-driven convergence stage with RLHF triggers. The result reduces latency-cost-accuracy trade-offs while supporting auditability and rollback for high-risk sectors such as health care and finance.
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。