Native Reasoning Models: Training Language Models to Reason on Unverifiable Data
Published in ICLR 2026 Poster, 2026
We propose a novel approach for training language models to reason on unverifiable data. The method enables native reasoning capabilities without requiring ground-truth supervision, addressing a fundamental challenge in training large language models to perform complex reasoning tasks.
Recommended citation: Wang, Y., Liu, Z., Li, X., Lu, C., & Yang, C. (2026). "Native Reasoning Models: Training Language Models to Reason on Unverifiable Data." ICLR 2026 Poster.
Download Paper