https://yanlunai.com/en/Recent content on YanlunAIHugo -- 0.152.2enMon, 24 Aug 2026 10:00:00 +0800https://yanlunai.com/en/topics/long-horizon-agents/Mon, 24 Aug 2026 10:00:00 +0800https://yanlunai.com/en/topics/long-horizon-agents/Tracking memory, state, tool use, recovery, and human intervention in long-running tasks.https://yanlunai.com/en/frontier/agent-memory-2026/Mon, 24 Aug 2026 09:00:00 +0800https://yanlunai.com/en/frontier/agent-memory-2026/The hard part is not storing history. It is deciding what to write, when to retrieve it, and when to forget.https://yanlunai.com/en/topics/ai-coding/Sun, 23 Aug 2026 10:00:00 +0800https://yanlunai.com/en/topics/ai-coding/How delegation, verification evidence, and expertise determine the real return from coding agents.https://yanlunai.com/en/industry/ai-coding-adoption/Sun, 23 Aug 2026 09:00:00 +0800https://yanlunai.com/en/industry/ai-coding-adoption/Teams now need to evaluate task boundaries, verification cost, and accountability—not only generation speed.https://yanlunai.com/en/topics/inference-routing/Sat, 22 Aug 2026 10:00:00 +0800https://yanlunai.com/en/topics/inference-routing/Moving beyond one benchmark to task quality, latency, throughput, and cost per successful task.https://yanlunai.com/en/labs/model-routing-lab/Sat, 22 Aug 2026 09:00:00 +0800https://yanlunai.com/en/labs/model-routing-lab/A reproducible routing skeleton that records task success, time to first token, total latency, and cost per successful task.https://yanlunai.com/en/topics/trustworthy-agents/Fri, 21 Aug 2026 10:00:00 +0800https://yanlunai.com/en/topics/trustworthy-agents/Designing for prompt injection, permission boundaries, tool side effects, and high-risk actions.https://yanlunai.com/en/posts/tools/2025-12-16-rag-toolchain-overview/Tue, 16 Dec 2025 09:30:00 +0800https://yanlunai.com/en/posts/tools/2025-12-16-rag-toolchain-overview/Treat RAG as a pipeline: data, chunking, indexing, retrieval, reranking, generation, and evaluation. Any stage can become the bottleneck.https://yanlunai.com/en/posts/papers/2025-12-15-agent-planning-view/Mon, 15 Dec 2025 20:10:00 +0800https://yanlunai.com/en/posts/papers/2025-12-15-agent-planning-view/An agent is more than a multi-turn chat interface: it maintains state, decomposes goals, uses tools, and recovers from failure.https://yanlunai.com/en/posts/papers/2025-12-15-clean-rag-engineering/Mon, 15 Dec 2025 17:40:00 +0800https://yanlunai.com/en/posts/papers/2025-12-15-clean-rag-engineering/Trustworthy RAG is not a prompt. It is a system for traceable evidence, grounded answers, and abstention under uncertainty.