Limitations/Comparisons and what is this model trained for?

#30
by ku7 - opened

I assumed by the name that this was primarily for python, but I had a rather interesting problem to solve.

It was in C++, and required basic understanding of RC and LC circuits.
I compared it against Bard, Claude, ChatGPT3.5, ChatCPT4. They each had their strengths and weaknesses. This could not do math whatsoever, and thus made bad assumptions, but once I did that for it, it produced the second-best code, only after Claude, and was the only one to spot an inefficiency I made in my code.
Did I get lucky with this test, or is it actually multi language?
What languages is it actually good at?

Sign up or log in to comment