Score: 1

Online Rubrics Elicitation from Pairwise Comparisons

Published: October 8, 2025 | arXiv ID: 2510.07284v1

By: MohammadHossein Rezaei , Robert Vacareanu , Zihao Wang and more

BigTech Affiliations: Scale AI

Potential Business Impact:

Teaches computers to write better answers by changing rules.

Business Areas:
Crowdsourcing Collaboration

Rubrics provide a flexible way to train LLMs on open-ended long-form answers where verifiable rewards are not applicable and human preferences provide coarse signals. Prior work shows that reinforcement learning with rubric-based rewards leads to consistent gains in LLM post-training. Most existing approaches rely on rubrics that remain static over the course of training. Such static rubrics, however, are vulnerable to reward-hacking type behaviors and fail to capture emergent desiderata that arise during training. We introduce Online Rubrics Elicitation (OnlineRubrics), a method that dynamically curates evaluation criteria in an online manner through pairwise comparisons of responses from current and reference policies. This online process enables continuous identification and mitigation of errors as training proceeds. Empirically, this approach yields consistent improvements of up to 8% over training exclusively with static rubrics across AlpacaEval, GPQA, ArenaHard as well as the validation sets of expert questions and rubrics. We qualitatively analyze the elicited criteria and identify prominent themes such as transparency, practicality, organization, and reasoning.

Country of Origin
🇺🇸 United States

Page Count
20 pages

Category
Computer Science:
Computation and Language