likelihood-ranking

Tag

Cards List
#likelihood-ranking

Likelihood Ranking doesn't Scale Like Prompting in LLMs

arXiv cs.CL ↗ · 2026-09-25 Cached

The paper investigates how likelihood ranking and prompting-based evaluation scale differently in LLMs, finding that likelihood accuracy remains stable across model scales while prompted performance improves, suggesting they probe distinct aspects of model behavior.

0 favorites 0 likes
← Back to home

Submit Feedback