Model

In this section, users can customize instructions, select an LLM model, Fall Back LLM model and adjust the temperature settings as desired.

• Instruction: The instructions field lets you personalize your Assistant's personality and style.
• Model: Choose various AI models based on your plan.
• Temperature: Adjust the temperature settings to control the assistant's creativity in responses (low being normal, high being more creative).

Guidelines to follow:

1. Adjusting the Assistant's Tone: Specify whether the Assistant should maintain a formal, casual, or emotionally specific tone. This choice will shape how it interacts in different contexts. For example, you might say: “You are a friendly and approachable AI Assistant.”

2. Addressing Different Types of Questions: Indicate whether the Assistant should deliver brief and direct answers or more comprehensive and thoughtful responses, depending on the nature of the question. It’s crucial to define how it should tackle factual inquiries versus opinion-based ones.

3. Managing Situations with No Available Answers: Clarify the Assistant's course of action when it lacks an answer. Should it courteously suggest the user look elsewhere, provide general guidance, or recommend alternative resources?.

4. Determining When to Include URLs: Specify the conditions under which the Assistant should provide links in its responses. Should it only share URLs when explicitly requested or should it proactively point out relevant resources?.

5. Focusing on Specific Areas of Expertise: If the Assistant needs to concentrate on particular subjects, you might guide it with a statement like: “You are an AI Assistant with a focus on environmental sustainability.” This helps narrow the Assistant's responses to a specific field.