User Guide
If you set a question to be public, anyone will be able to view them. Please do not put any sensitive info (such as API keys) in public questions.
You can always go back to your previous question history in My Threads and change the visibility settings on your questions.
How to ask questions
To get the most accurate and relevant responses from Adrenaline's LLMs, there are some common pitfalls you should avoid. Providing sufficient context and specificity in your queries will help Adrenaline better understand your needs and provide more accurate and helpful answers.
Use Descriptive Language
When posing your question, use descriptive language and provide as much context as possible. Instead of asking vague or ambiguous questions, aim to be specific and clearly convey the topic or domain you're inquiring about.
Provide Relevant Details
Many questions involve code snippets.
Avoid pasting them directly into the query field. Instead, utilize the Snippet+ feature for improved readability and easier debugging.
Including relevant details in your question can significantly improve the quality of the response you receive. The more context and information you provide, the better the LLM can understand your query and provide an accurate and helpful answer.
File/Directory Path:
When referencing specific files or directories, include the full path for better contextual understanding.
Project Details:
If your question pertains to a particular project, share details about its structure, dependencies, or any other relevant information.
Desired Functionality:
Describe the outcome or behavior you're aiming to achieve for questions about specific features or functionalities.
- Example: "I'm working on a data processing pipeline in my 'data_analysis' project. The 'clean_data.py' script isn't removing duplicate entries as expected. Can you help me troubleshoot the code?"
Adjust Your Phrasing
On rare occasions, specific phrasings can lead to irrelevant web results that hinder the Adrenaline's final answer quality.
If the LLM relies on web results unrelated to your topic, try rephrasing your question in a way you'd typically use a search engine.