Context Stuffing

Definitions

1. The practice of adding too much material to a prompt or context window in the hope that more information will improve the answer, often with diminishing returns or worse performance. Context stuffing can increase cost, latency, and confusion for the model.

   In plain English: Packing too much information into an AI prompt or context window.

   Example: "The assistant improved after they stopped context stuffing and retrieved only the few policy pages actually relevant to the user's question."

Related Terms

• AI Index
• Chain of Density
• Chain of Verification
• Few-Shot
• RAG Retriever
• Relevance Score
• Retrieval Model
• Retrieval Pipeline
• Retriever
• Soft Prompt
• Unlocking
• Zero-Shot