Excellent summary. I really like the suggestions of using the tags to identify the sections, and keeping a library. I also keep notes on the successes and failures of my prompts, and how they work in different LLMs.
Nothing too systematic, it's just that some prompts seem to drive a lot of hallucination (or ignoring certain parts of the prompts) ... so it's something in my prompts, not necessarily the LLMs. I need to get better at understanding how to do validation. I'd love more advice/help on that.
Excellent summary. I really like the suggestions of using the tags to identify the sections, and keeping a library. I also keep notes on the successes and failures of my prompts, and how they work in different LLMs.
Jed - thanks for the feedback, what differences have you noticed between the different LLMs?
Nothing too systematic, it's just that some prompts seem to drive a lot of hallucination (or ignoring certain parts of the prompts) ... so it's something in my prompts, not necessarily the LLMs. I need to get better at understanding how to do validation. I'd love more advice/help on that.