|
|
|
|
|
|
by ComputerGuru
144 days ago
|
|
|
You can’t ask GPT to assess the situation. That’s not the kind of question you can count on a an LLM to accurately answer. Playing with the system prompts, temperature, and max token output dials absolutely lets you make enough headway (with the 5 series) in this regard to demonstrably render its self-analysis incorrect. |
|
|