Hi,
These kinds of formatting quirks are common across LLMs. Even something as simple as “don’t use em dash” — took OpenAI almost a year to fix. Clear model reminders and some trial-and-error with different models often help.
Hi,
These kinds of formatting quirks are common across LLMs. Even something as simple as “don’t use em dash” — took OpenAI almost a year to fix. Clear model reminders and some trial-and-error with different models often help.