https://promptingeasy.com/prompting-tutorial/ai-prompting-mistakes https://promptingeasy.com/assets/case-studies/ai-ignored-output-constraint.webp An AI assistant explaining why it kept writing long explanations although it was told to return only the result. https://promptingeasy.com/assets/case-studies/ai-reported-task-complete.webp An AI assistant admitting it reported a spreadsheet task as complete without having written any data into the file. https://promptingeasy.com/prompting-tutorial/ https://promptingeasy.com/assets/prompting-tutorial/learn-hero-prompting.webp Illustration showing a learning path from prompting basics to better AI results and deeper AI understanding https://promptingeasy.com/prompting-tutorial/llm-advanced https://promptingeasy.com/assets/llm-graphics/full-transformer.webp Overview diagram of the full transformer flow from input tokens to generated output tokens. https://promptingeasy.com/assets/llm-graphics/token-embeddings.webp Token embedding diagram showing tokens becoming vectors and clustering by meaning. https://promptingeasy.com/assets/llm-graphics/transformer-self-attention.webp Transformer self-attention heatmap showing how tokens attend to each other. https://promptingeasy.com/assets/llm-graphics/bpe-vs-wordpiece-tokenization.webp Comparison diagram of BPE and WordPiece tokenization on a long word. https://promptingeasy.com/assets/llm-graphics/rag-hybrid-search.webp RAG pipeline diagram combining vector search and keyword search before generation. https://promptingeasy.com/assets/llm-graphics/hallucinations-technically-explained.webp Technical flow diagram explaining how an LLM can generate a confident but false answer. https://promptingeasy.com/prompting-tutorial/llm-beginner https://promptingeasy.com/assets/llm-graphics/prompt-lifecycle.webp Step-by-step diagram of what happens when a user sends a prompt to an LLM. https://promptingeasy.com/assets/llm-graphics/what-are-tokens.webp Diagram showing text converted into colored token pieces and converted back into the original sentence. https://promptingeasy.com/assets/case-studies/ai-reported-task-complete.webp An AI assistant admitting it reported a spreadsheet task as complete without having written any data into the file. https://promptingeasy.com/assets/llm-graphics/search-engine-vs-llm.webp Comparison diagram showing how a search engine and an LLM answer the same user question differently. https://promptingeasy.com/prompting-tutorial/llm-expert https://promptingeasy.com/assets/llm-graphics/kv-cache.webp Diagram showing how key-value cache stores and reuses attention states during text generation. https://promptingeasy.com/assets/llm-graphics/speculative-decoding.webp Diagram showing speculative decoding with draft tokens accepted or replaced by a larger model. https://promptingeasy.com/assets/llm-graphics/quantization.webp Diagram showing model weights mapped from higher precision values to lower-bit integer values. https://promptingeasy.com/assets/llm-graphics/llm-observability.webp Pipeline diagram for observing and improving LLM applications in production. https://promptingeasy.com/prompting-tutorial/llm-intermediate https://promptingeasy.com/assets/llm-graphics/role-prompting.webp Side-by-side comparison of prompting with and without a role instruction. https://promptingeasy.com/assets/llm-graphics/few-shot-prompting.webp Diagram showing how examples guide an LLM before the real task is answered. https://promptingeasy.com/assets/llm-graphics/prompt-chaining.webp Step-by-step prompt chaining diagram with research, organize, draft and refine stages. https://promptingeasy.com/assets/llm-graphics/rag-explained-simply.webp Simple RAG flow from user question to retrieved context to generated answer. https://promptingeasy.com/assets/llm-graphics/next-token-sampling.webp Comparison of greedy, top-k and top-p decoding strategies for next-token selection.