Increased specificity refers to the prompt engineering technique of providing more detailed, granular instructions to a language model by increasing the specificity of each step. For example: Original: Remove names from the text. More specific: 1. Identify all first and last names in the text. 2. Replace each first name with [FIRST NAME]. 3. Replace each last name with [LAST NAME]. The more specific steps make it clear exactly what constitutes a name and how each name should be handled. This leaves less room for interpretation by the language model compared to vague, high-level instruction like "remove names". Increased specificity reduces ambiguity, makes the expected output clearer, and minimizes the chances of the model misinterpreting or improperly executing the requested task. It is especially important for complex tasks comprising multiple sub-steps. More specific prompting leads to more accurate results. [[summarization]] < [[Hands-on LLMs]]/[[5 Prompting]] > [[role prompts]]