Let's try a new strategy.  Given the file axioms.csv attached, verify that it has a column showing the FOL formula (fol) and a column (english) describe the intention of the formula in english. The try to extract concepts and relation directly from fol formulas was frustated. So I want to experiment to extract concepts and formulas directly from the english sentence, using the capabilities of a LLM model. I have access to LLama3.1-8B. Create a pipeline using python that can read the csv file and generate a tentative list of concepts and relations. Analize also existing patterns on the english descriptions that could be trained to improve the results.

Given the file axioms.csv attached, verify that it has a column showing the FOL formula (fol) and a column (english) describe the intention of the formula in english.
Try to update this csv with a new column (english2) with a english sentence that translate more precisaly what the FOL formula is expressig.
