Using Custom Instructions to enhance academic output quality in ChatGPT | Alpha | PandaiTech

Using Custom Instructions to enhance academic output quality in ChatGPT

Techniques for using Custom Instructions to 'prime' the model with your professional background (e.g., a PhD researcher) to get more in-depth, technical answers rather than generic ones.

Learning Timeline
Key Insights

The 'Steering' Technique for Output Quality

By default, ChatGPT provides generalized responses. By applying 'Identity Steering' within Custom Instructions, the AI shifts its response parameters from a layperson's level to that of a subject matter expert.

Impact on Discovering Breakthroughs

When you 'prime' the model with highly specific context, the AI is more likely to connect rare technical data points and can even suggest solutions comparable to current laboratory research.
Prompts

Expert Persona Priming for Researchers

Target: ChatGPT
I am a frontier researcher in a [State Your Lab/Field] using [Specific Tools/Science]. Respond to all my queries at a high academic level suitable for a PhD-level specialist. Do not provide undergraduate-level summaries; focus on technical depth and recent breakthroughs in the field.
Step by Step

How to Configure Custom Instructions for Academic Output

  1. Open ChatGPT and click on your profile or username in the bottom-left corner of the screen.
  2. Select the 'Customize ChatGPT' or 'Custom Instructions' menu from the list of options.
  3. In the first input box ('What would you like ChatGPT to know about you...'), enter details regarding your specific expertise, such as your research position, PhD field, or the lab where you work.
  4. Specify the types of tools, methodologies, or specific sciences you use in your daily tasks.
  5. In the second input box ('How would you like ChatGPT to respond?'), set the response level to 'Academic Frontier Level' or 'Expert Researcher'.
  6. Instruct the model to avoid undergraduate-level responses and focus instead on cutting-edge findings.
  7. Click the 'Save' button to activate these settings for all future conversations.