Saved in:
Bibliographic Details
Main Authors: Zhang, Yichi, Zhang, Siyuan, Huang, Yao, Xia, Zeyu, Fang, Zhengwei, Yang, Xiao, Duan, Ranjie, Yan, Dong, Dong, Yinpeng, Zhu, Jun
Format: Preprint
Published: 2025
Subjects:
Online Access:https://arxiv.org/abs/2502.02384
Tags: Add Tag
No Tags, Be the first to tag this record!
Table of Contents:
  • Ensuring the safety and harmlessness of Large Language Models (LLMs) has become equally critical as their performance in applications. However, existing safety alignment methods typically suffer from safety-performance trade-offs and the susceptibility to jailbreak attacks, primarily due to their reliance on direct refusals for malicious queries. In this paper, we propose STAIR, a novel framework that integrates SafeTy Alignment with Itrospective Reasoning. We enable LLMs to identify safety risks through step-by-step analysis by self-improving chain-of-thought (CoT) reasoning with safety awareness. STAIR first equips the model with a structured reasoning capability and then advances safety alignment via iterative preference optimization on step-level reasoning data generated using our newly proposed Safety-Informed Monte Carlo Tree Search (SI-MCTS). We further train a process reward model on this data to guide test-time searches for improved responses. Extensive experiments show that STAIR effectively mitigates harmful outputs while better preserving helpfulness, compared to instinctive alignment strategies. With test-time scaling, STAIR achieves a safety performance comparable to Claude-3.5 against popular jailbreak attacks. Relevant resources in this work are available at https://github.com/thu-ml/STAIR.