If you give an LLM a list of items, and then ask it to order by some property (alphabetic ordering, weight, date or birth, etc.) there will be certain orderings that succeed, certain ones that fail, and certain ones that fail abysmally. Can we use this to figure out what features LLMs have easier access to?
If you are inspired by this idea, you can reach out to the authors for collaboration or cite it:
@misc{holtzman-llm-onepass-orderings-2026,
author = {Holtzman, Ari},
title = {LLM One-pass Orderings},
year = {2026},
url = {https://hypogenic.ai/ideahub/idea/SBpCYqmYPvNkXPFoJdrK}
}Please sign in to comment on this idea.
No comments yet. Be the first to share your thoughts!