Is it possible to use Chat GPT to create a Periodic Table of Elements that provides the maximum amount of accurate element data based on the arrangement of the Elements and the tables numbering system?
It’s possible but I am not sure if a LLM is the right tool for the job.
This is because language models excel at natural language and for precision, for example when building a solution to attach several numbers to items, you will likely often end-up with a mistake here and there. This is because the model is stochastic by design.
The often suggested solution is then to provide the table to the model and have it repeat the contents back to the user.
But then again, why use a language model for this task?
You would likely be better off having the model comment or explain the table to the user using natural language.