Skip to main content

49 posts tagged with "llm"

View All Tags

AI Daily Digest: DeepSeek-V4 发布百万上下文 Agent 专用模型,Google Deep Research Max 重新定义自主研究 - 2026/04/24

· 5 min read
Yi Wang
Full Stack & AI Engineer

今日 AI 行业有两个重磅发布。DeepSeek 发布 V4 系列,以 Hybrid Attention 架构实现百万级上下文窗口在 Agent 场景下的可用性突破;Google 推出基于 Gemini 3.1 Pro 的 Deep Research Max,支持 MCP 协议接入和异步企业工作流。学术方面,arXiv 新增多篇关于 Agentic AI 和高效微调的论文。

AI Daily Digest: Google Cloud Next '26 全面拥抱 Agent,Anthropic Glasswing 重新定义 AI 安全 - 2026/04/23

· 7 min read
Yi Wang
Full Stack & AI Engineer

今天 AI 行业迎来密集发布。Google Cloud Next '26 大会以 "Agentic Cloud" 为主题,全面转向 Agent 架构;Anthropic 联合科技巨头发起 Project Glasswing 安全联盟;Kubernetes v1.36 正式发布;Docker Hub 再遭供应链攻击。以下是今日要闻。

AI Daily Digest: GLM-5.1 开源击败 GPT-5.4,推理时自动纠错新突破 - 2026/04/21

· 7 min read
Yi Wang
Full Stack & AI Engineer

开源模型 GLM-5.1 在 SWE-Bench Pro 上超越 GPT-5.4,宣告"开源落后闭源 6 个月"的叙事终结。与此同时,arXiv 本周论文聚焦推理时自动纠错(KV-Cache 回滚)和 Agent 预测系统,Google 开放 AI Studio 订阅者深度使用权限,终端优先 AI 开发工具成为新范式。

AI Daily Digest: Agent 成功率暴增 12%→66%,RL 奖励作弊检测新方法 - 2026/04/20

· 6 min read
Yi Wang
Full Stack & AI Engineer

Stanford 2026 AI Index 发布最新数据:AI Agent 任务成功率从去年的 12% 跃升至 66%,AI Agent 相关网络流量暴增 7,851%。与此同时,arXiv 本周论文聚焦 AI 安全审计和 RL 奖励作弊检测,Google 发布机器人领域新模型,Docker 公开其 Agent 沙箱架构。

Context Engineering: The Strategic RAM of AI

· 9 min read
Yi Wang
Full Stack & AI Engineer

In the early days of the Generative AI revolution, the industry was obsessed with "Parameters." We measured progress by the billions, then trillions, of weights packed into a model's neural architecture. But by 2026, the consensus has shifted. As we stand in the era of Gemini 3.0 and Claude 4, we’ve realized that raw intelligence is useless without a high-fidelity, low-latency "Working Memory."

Welcome to the age of Context Engineering. If the LLM is the CPU, context is the RAM. And just as in traditional computing, the way we manage this RAM defines the ceiling of what the system can actually accomplish.