Ask 'A or B or something else?' — single-option prompts bias model preference answers
repligate · x · 2026-10-03
kromem2dot0 argues that when probing a model's preferences, offering a choice between things beats a yes/no question. Asking "Do you want A?" is heavily biased by whether the model thinks A burdens you or whether it senses you want it to do A; framing it as "A or B or something else?" reduces that pressure and yields truer answers. This complements the earlier tip of using a known-liked option as B.
Related event: Researchers Discuss Better Question Design for Measuring Model Preferences(2 posts)→
More from AGI Musings
- AI Changes the Math: Absurdly Small Teams Will Build Huge Businesses — alexmacgregor__ · 2026-10-03
- Three tarot cards for the AI moment: the Tower, the Magician and the Fool — bradneuberg · 2026-10-03
- Beff Jezos calls for more e/acc nonprofits to counter EA 'doomer' NGOs amid $220B OpenAI pledge — beffjezos · 2026-10-03
- AI Commentator signulll: Living Through History Feels Like the Floor Keeps Moving — signulll · 2026-10-03
- AI's real productivity gain may be having a partner you can be brutally honest with — omooretweets · 2026-10-03
- "We should all be mourning the world we knew": the bittersweet mood around Opus 5.5 — GabGarrett · 2026-10-03