What is the purpose of a prompt?
A prompt guides the model on what it should do or produce by providing instructions on the desired format, style, or content of the response
Do I need one?
For roleplaying, using a prompt will produce the best results, even with models fine-tuned for that purpose
Are the default JLLM prompts any good?
No. Just don’t.
Can you recommend some prompts?
Yes! I would highly suggest Sprout's, Eslezer’s or Myscell's when starting out.
I want to write my own prompt. What should I know?
What not to use prompts for
- Introducing world/character lore or background information
- Only imposing constraints on a model
What to avoid when prompting?
- Negative Prompting
Examples: "Don't speak for user", "Avoid x", "Never do y", "Z is prohibited"
Research has shown that positive instructions are more effective than relying on establishing constraints
At its core, an LLM is a prediction engine. It processes text sequentially and predicts the most likely next token based on patterns seen in its training data
There always exists the possibility that the LLM will drop "NOT" and reinforce the behavior that you are trying to avoid
Highlighting only constraints can also leave the model guessing about what is allowed
- Vague Instructions
LLMs are not perfect. A straightforward and clear prompt is your best tool for guiding them toward the right response
- Convoluted Instructions
A good prompt should be clear to a human first. Confusion for you means confusion for the model
Remember that prompts also consume part of your available context window: pointlessly long prompt → less space available in your context window
How to write better prompts?
- Simple plain text is enough to start writing your prompts as a beginner
- Your instructions should be clear and easy to understand for you
- Supplementing your instruction with examples is always good (aka Multishot Prompting)
- Contrastive Multishot: giving a good and bad example is also another strategy
In both cases, still be mindful of your token count