https://guijiagi.com/tags/llm/Recent content in LLM on 硅基 AGI · 智能体学习与测评Hugozh-cn本站内容采用 CC BY-NC-SA 4.0 国际许可协议授权Thu, 02 Jul 2026 11:31:00 +0800https://guijiagi.com/posts/llamaindex-2026-guide/Thu, 02 Jul 2026 11:31:00 +0800https://guijiagi.com/posts/llamaindex-2026-guide/2026年LlamaIndex框架使用指南，构建数据驱动的LLM应用https://guijiagi.com/posts/ollama-2026-ecosystem/Thu, 02 Jul 2026 11:25:00 +0800https://guijiagi.com/posts/ollama-2026-ecosystem/2026年Ollama生态系统全面解析，从模型管理到生产部署https://guijiagi.com/posts/pruning-llm-2026/Thu, 02 Jul 2026 11:01:00 +0800https://guijiagi.com/posts/pruning-llm-2026/从幅度剪枝到稀疏训练，系统对比LLM剪枝的两大路线及2026年最新方案https://guijiagi.com/posts/distillation-llm-2026/Thu, 02 Jul 2026 11:00:00 +0800https://guijiagi.com/posts/distillation-llm-2026/从Logit蒸馏到特征对齐，全面梳理大模型知识蒸馏的方法与实践https://guijiagi.com/posts/specul-ive-decoding-2026/Thu, 02 Jul 2026 10:55:00 +0800https://guijiagi.com/posts/specul-ive-decoding-2026/从Draft-Verify到Medusa头，系统梳理推测解码的演进与2026年最新方案https://guijiagi.com/posts/kv-cache-optimization-techniques/Thu, 02 Jul 2026 10:53:00 +0800https://guijiagi.com/posts/kv-cache-optimization-techniques/从PagedAttention到量化缓存，系统梳理KV Cache的各类优化方案https://guijiagi.com/posts/llm-benchmark-2026-july-comparison/Thu, 02 Jul 2026 10:50:00 +0800https://guijiagi.com/posts/llm-benchmark-2026-july-comparison/2026年7月主流大语言模型在多项基准测试上的全面对比分析https://guijiagi.com/posts/ai-common-sense-problem/Thu, 02 Jul 2026 10:41:00 +0800https://guijiagi.com/posts/ai-common-sense-problem/为什么拥有万亿参数的LLM仍然缺乏基本常识？https://guijiagi.com/posts/embodied-ai-progress-2026/Thu, 02 Jul 2026 10:36:00 +0800https://guijiagi.com/posts/embodied-ai-progress-2026/2026年具身智能重大进展：大语言模型驱动的智能机器人https://guijiagi.com/posts/differential-privacy-llm/Thu, 02 Jul 2026 10:26:00 +0800https://guijiagi.com/posts/differential-privacy-llm/深入探讨差分隐私技术在大语言模型训练和推理中的应用方法与最新进展https://guijiagi.com/posts/prompt-engineering-2026/Tue, 30 Jun 2026 09:10:00 +0800https://guijiagi.com/posts/prompt-engineering-2026/2026年Prompt工程全景指南：从基础技巧到结构化Prompt、多模态Prompt、Agent Prompt、企业级Prompt管理系统https://guijiagi.com/posts/ai-translation-2026-comparison/Sun, 28 Jun 2026 12:10:00 +0800https://guijiagi.com/posts/ai-translation-2026-comparison/2026年AI翻译工具深度对比，涵盖DeepL、Google Translate与LLM在专业翻译场景的表现https://guijiagi.com/posts/ai-security-operations-llm-soc/Sun, 28 Jun 2026 12:06:00 +0800https://guijiagi.com/posts/ai-security-operations-llm-soc/2026年用LLM增强安全运营中心(SOC)的实践指南，涵盖告警分诊、威胁狩猎与事件响应https://guijiagi.com/posts/llm-hallucination-2026-analysis/Sun, 28 Jun 2026 11:10:00 +0800https://guijiagi.com/posts/llm-hallucination-2026-analysis/LLM幻觉问题的深度根因分析与2026年最新缓解策略：从RAG到自验证的完整方案https://guijiagi.com/posts/lm-studio-2026-review/Sun, 28 Jun 2026 11:00:00 +0800https://guijiagi.com/posts/lm-studio-2026-review/深度评测 LM Studio 2026 版本，从模型管理、对话界面、API 服务到 RAG 构建的全面分析https://guijiagi.com/posts/ollama-2026-local-llm/Sun, 28 Jun 2026 11:00:00 +0800https://guijiagi.com/posts/ollama-2026-local-llm/全面介绍 Ollama 2026 版本的特性、模型管理、API 使用、性能优化与生产部署最佳实践https://guijiagi.com/posts/vllm-2026-deployment-guide/Sun, 28 Jun 2026 11:00:00 +0800https://guijiagi.com/posts/vllm-2026-deployment-guide/从单机部署到分布式集群，全面覆盖 vLLM 2026 的安装、配置、优化与生产运维实践https://guijiagi.com/posts/prompt-compression-techniques/Sun, 28 Jun 2026 10:45:00 +0800https://guijiagi.com/posts/prompt-compression-techniques/2026年Prompt压缩技术全景：从符号化压缩到语义压缩的最新方法与工程实现https://guijiagi.com/posts/agent-observability-otel-2026/Sun, 28 Jun 2026 10:10:00 +0800https://guijiagi.com/posts/agent-observability-otel-2026/如何使用OpenTelemetry标准构建LLM Agent的全链路可观测性体系，涵盖Traces、Metrics、Logs三大支柱https://guijiagi.com/posts/prompt-engineering-2026-practices/Sun, 28 Jun 2026 10:00:00 +0800https://guijiagi.com/posts/prompt-engineering-2026-practices/2026年Prompt工程已从个人技巧发展为系统化工程体系，本文全面梳理从设计、测试、版本管理到监控的完整流程https://guijiagi.com/posts/llm-evaluation-pipeline-benchmark-to-custom/Sun, 28 Jun 2026 10:00:00 +0800https://guijiagi.com/posts/llm-evaluation-pipeline-benchmark-to-custom/系统讲解大模型评估流水线的搭建方法，涵盖主流 Benchmark 集成、自定义评测设计和自动化 CI/CDhttps://guijiagi.com/posts/llm-code-gen-eval/Sat, 27 Jun 2026 15:00:00 +0800https://guijiagi.com/posts/llm-code-gen-eval/LLM代码生成能力评测https://guijiagi.com/posts/llm-benchmark-credibility/Sat, 27 Jun 2026 15:00:00 +0800https://guijiagi.com/posts/llm-benchmark-credibility/LLM评测榜单可信度分析https://guijiagi.com/posts/llm-scaling-wall/Sat, 27 Jun 2026 15:00:00 +0800https://guijiagi.com/posts/llm-scaling-wall/大模型Scaling撞墙了吗https://guijiagi.com/posts/llm-output-filter-design/Sat, 27 Jun 2026 15:00:00 +0800https://guijiagi.com/posts/llm-output-filter-design/大模型输出过滤系统设计https://guijiagi.com/posts/llm-inference-pipeline/Sat, 27 Jun 2026 15:00:00 +0800https://guijiagi.com/posts/llm-inference-pipeline/大模型推理流水线设计https://guijiagi.com/posts/llm-inference-gateway/Sat, 27 Jun 2026 15:00:00 +0800https://guijiagi.com/posts/llm-inference-gateway/大模型推理网关架构https://guijiagi.com/posts/function-calling-internals/Sat, 27 Jun 2026 15:00:00 +0800https://guijiagi.com/posts/function-calling-internals/函数调用Function Calling底层原理https://guijiagi.com/posts/long-context-evolution/Sat, 27 Jun 2026 15:00:00 +0800https://guijiagi.com/posts/long-context-evolution/长上下文窗口技术演进路线https://guijiagi.com/posts/ai-testing-strategy/Thu, 25 Jun 2026 12:00:00 +0800https://guijiagi.com/posts/ai-testing-strategy/传统软件测试方法在 AI 应用中不够用。本文构建从单元测试到红队测试的完整 AI 测试金字塔，覆盖确定性、概率性和对抗性测试。https://guijiagi.com/posts/a-b-testing-for-llm/Thu, 25 Jun 2026 12:00:00 +0800https://guijiagi.com/posts/a-b-testing-for-llm/系统介绍 LLM 产品的 A/B 测试方法，涵盖实验设计、样本量计算、统计显著性检验、业务指标选择与常见陷阱。https://guijiagi.com/posts/llm-deployment-k8s/Thu, 25 Jun 2026 12:00:00 +0800https://guijiagi.com/posts/llm-deployment-k8s/从 GPU 节点池配置、模型服务部署、自动扩缩容到推理优化，完整介绍 LLM 在 Kubernetes 上的生产级部署方案。