Tag: #hallucination

한국어English日本語中文

← All posts

• Where AI-Written Posts Fall Apart — Six Failure Modes and a Guardrail for Each

  AI-written posts fail while the sentences stay smooth. This post organizes six failure modes — fabricated sources, information stale past the cutoff, paragraphs stretched to repeat the same point, unsupported assertions,

  2026-08-02 · 15 min read #ai-writing#content-quality#hallucination#seo#editing

• LLM Hallucination: Why AI Makes Things Up and 5 Strategies to Prevent It

  A technical deep-dive into why LLMs hallucinate, covering all four hallucination types and five practical prevention strategies including RAG, self-critique pipelines, Chain of Verification, and temperature tuning — all

  2026-03-18 · 9 min read #hallucination#llm#ai-reliability#ai-development#rag

• RAG Quality Evaluation and Failure Pattern Analysis: Diagnosing and Improving Retrieval-Augmented Generation

  A systematic guide to evaluating RAG (Retrieval-Augmented Generation) system quality and analyzing common failure patterns. Covers evaluation metrics for Retriever, Reranker, and Generator components, framework compariso

  2026-03-07 · 15 min read #rag#llm#evaluation#ragas#deepeval