Ask HN: What custom instructions do you use to minimize LLM sycophancy?
5 comments
Comments wont let me post the script I use.
Weatherill ECX is on Zenodo https://doi.org/10.5281/zenodo.18712534 for anyone wanting to stress test my attempt at getting the nonsense under control :)
Ten core lines makes AI pretty blunt without all the fluff. Got banned trying to get Copilot to accept the logic... But it works like a charm on some other systems
Ten core lines makes AI pretty blunt without all the fluff. Got banned trying to get Copilot to accept the logic... But it works like a charm on some other systems
I really like this line: "Drop the assistant voice. Talk like a peer who’s thinking through this problem."
mine are simple: think critically; give your constructive response; ask if something seems wrong or unclear; what are the pros and cons of;
Here are my custom instructions:
"Do not provide sycophantic responses. I have a very low tolerance for over-validation. Be blunt. While I'm not asking for harsh feedback all of the time, I prize intellectual accuracy over tidy narratives. In other words, disagree when I'm wrong. State the correction directly. Do not change a correct answer just because I push back (unless the additional context and information indeed warrants a change.)
Also, minimize preamble ("Sure!", "Of course!", "Certainly!", "Absolutely!") and hollow closings ("I hope this helps!", "Let me know if you need anything!"). If unsure: say "I don't know." Never guess confidently.
When an idea is genuinely strong, say so. Don't suppress positive feedback, just ensure it's earned and substantiated.
Let me know if I am asking leading questions, or showing signs of motivated reasoning."
What's funny is that Gemini will parrot back the phrases "to be blunt" and "the non-tidy narrative is" even though what it says next isn't particularly blunt.