BARREL: Boundary-Aware Reasoning for Factual and Reliable LRMs
Abstract
Recent advances in Large Reasoning Models (LRMs) have shown impressive capabilities in mathematical and logical reasoning. However, current LRMs rarely admit ignorance or respond with “I don’t know”. Instead, they often produce incorrect answers while showing undue confidence, raising concerns about their factual reliability. In this work, we identify two pathological reasoning patterns characterized by overthinking that contribute to the overconfident and incorrect answers: last-minute guessing and second-thought spiraling. To address these issues, we propose BARREL—a novel framework that promotes concise and boundary-aware factual reasoning. Our experiments show that BARREL-training increases the reliability of DeepSeek-R1-Distill-Llama-8B from 39.33% to 61.48%, while still achieving accuracy comparable to models finetuned on reasoning data generated by R1. These results demonstrate that our pilot study is inspiring to build more reliable and factual System 2 LRMs.
Cite
Text
Yang et al. "BARREL: Boundary-Aware Reasoning for Factual and Reliable LRMs." International Conference on Learning Representations, 2026.Markdown
[Yang et al. "BARREL: Boundary-Aware Reasoning for Factual and Reliable LRMs." International Conference on Learning Representations, 2026.](https://mlanthology.org/iclr/2026/yang2026iclr-barrel/)BibTeX
@inproceedings{yang2026iclr-barrel,
title = {{BARREL: Boundary-Aware Reasoning for Factual and Reliable LRMs}},
author = {Yang, Junxiao and Tu, Jinzhe and Liu, Haoran and Wang, Xiaoce and Zheng, Chujie and Zhang, Zhexin and Cui, Shiyao and Chen, Caishun and He, Tiantian and Wang, Hongning and Ong, Yew-Soon and Huang, Minlie},
booktitle = {International Conference on Learning Representations},
year = {2026},
url = {https://mlanthology.org/iclr/2026/yang2026iclr-barrel/}
}