UniRRM: Unified Reasoning Reward Models Across Languages and Evaluation Paradigms

Published in ICML, 2026

Abstract: