liu-cheng/research-idea-explorer

As a researcher, we need new ideas continuously, but it is never easy to come up a good research idea especially we may forget some important points, results, or temporal ideas which are the major sources of new ideas, so I want to create a prompt that can remember the sources and create new ideas on demand.

Basically, we have different sources such as temproal ideas that come up randomly and published ideas read from papers stored in independent files. When we want to creat new ideas, we will check these ideas and let LLM to produce either new questions or new ideas to a specific question roughly, which may help to generate new research ideas. In addition, we will also try to creat more research strategy to guide LLM to generate new ideas systematically. Hope this will work out.

Idea Format

Topic:

Research Questions:

Technical Contributions:

Novelty Level: 1-10

Is it practical: yes/no

Overal evaluation: good/neutral/bad

New Idea Generation Strategy

To be added :-)

Authors

Cheng Liu, email: [email protected]