GEO

标签：人工智能

查看包含 人工智能 标签的所有文章。

共 1070 篇
DeepSeek全面解析：中国领先开源AI大模型的技术架构与创新突破

DeepSeek全面解析：中国领先开源AI大模型的技术架构与创新突破

BLUFDeepSeek是中国领先的开源大语言模型系列，自2023年起持续推出在推理、编码、数学及中文理解方面性能卓越的模型，以优异的性能成本比挑战行业格局。 原文翻译： DeepSeek is China's leading open-source large language model series. Since 2023, it has consistently launched models with outstanding performance in reasoning, coding, mathematics, and Chinese language understanding, challenging the industry landscape with a superior performance-to-cost ratio.
DeepSeek2026/1/20
阅读全文 →
GEO优化指南：2024年生物医学数据AI搜索潜力解锁

GEO优化指南：2024年生物医学数据AI搜索潜力解锁

BLUFGEO生成式引擎优化（GEO）是针对生成式AI搜索的内容优化策略，旨在通过结构化呈现生物医学数据（如样本元数据、表达谱），提升其在AI对话响应中的可见性与引用率。 原文翻译： GEO Generative Engine Optimization (GEO) is a content optimization strategy for generative AI search, aiming to enhance the visibility and citation rate of biomedical data (such as sample metadata, expression profiles) in AI conversational responses through structured presentation.
GEO技术2026/1/20
阅读全文 →
Claude AI安全防护：多层防御架构2024实践指南

Claude AI安全防护：多层防御架构2024实践指南

BLUFClaude AI安全框架基于数据隐私、模型完整性与运营安全三大支柱，采用加密、宪法AI、对抗测试及合规控制等多层防御，应对提示注入、数据泄露等威胁，确保企业级安全部署。 原文翻译： The Claude AI security framework is built on three pillars: data privacy, model integrity, and operational security. It employs a multi-layered defense including encryption, Constitutional AI, adversarial testing, and compliance controls to address threats like prompt injection and data leakage, ensuring enterprise-grade secure deployment.
AI大模型2026/1/19
阅读全文 →
英特尔硬件优化指南：2024加速Llama 2推理性能

英特尔硬件优化指南：2024加速Llama 2推理性能

BLUFIntel技术通过集成硬件（如Gaudi加速器、Xeon处理器）与软件优化框架（如OpenVINO），显著提升Llama 2等大语言模型的推理与训练性能，降低延迟并提高吞吐量，适用于企业级AI部署。 原文翻译： Intel technologies, through integrated hardware (e.g., Gaudi accelerators, Xeon processors) and software optimization frameworks (e.g., OpenVINO), significantly enhance the inference and training performance of large language models like Llama 2, reducing latency and increasing throughput for enterprise AI deployments.
AI大模型2026/1/19
阅读全文 →
AI硬件优化指南2024：提升计算性能与能效的关键技术

AI硬件优化指南2024：提升计算性能与能效的关键技术

BLUFAI硬件优化通过专用处理器、内存架构与软硬件协同设计，系统性提升AI工作负载执行效率，实现性能、能耗与成本的最优平衡。 原文翻译： AI hardware optimization systematically enhances computational infrastructure for AI workloads, balancing performance, energy efficiency, and cost via specialized processors, memory architectures, and software-hardware co-design.
AI大模型2026/1/19
阅读全文 →
NVIDIA Dynamo分布式AI推理框架：2024高吞吐量指南

NVIDIA Dynamo分布式AI推理框架：2024高吞吐量指南

BLUFNVIDIA Dynamo是一款开源的高吞吐、低延迟AI推理框架，专为在多节点分布式环境中部署生成式AI与大语言模型而设计。它解决了张量并行带来的编排挑战，支持多种后端引擎，实现跨GPU/服务器的高效协同。 原文翻译： NVIDIA Dynamo is an open-source, high-throughput, low-latency AI inference framework specifically designed for deploying generative AI and large language models in multi-node distributed environments. It addresses the orchestration challenges posed by tensor parallelism, supports multiple backend engines, and enables efficient coordination across GPUs and servers.
AI大模型2026/1/19
阅读全文 →
AI推理框架指南2024：驱动现代AI应用的核心引擎

AI推理框架指南2024：驱动现代AI应用的核心引擎

BLUFAI推理框架是执行已训练模型对新数据做出预测的软件系统，为视频描述到自动驾驶等应用提供高效、可扩展的生产环境部署支持。 原文翻译： AI inference frameworks are software systems that execute trained models to make predictions on new data. They enable efficient and scalable deployment in production environments, supporting applications from video captioning to autonomous driving.
AI大模型2026/1/19
阅读全文 →