← 研踪 PaperTrailACL Anthology · ACL 2026:acl:2026.acl-long.1570 · 记录 #1
Resolving the Security-Auditability Dilemma with Auditable Latent Chain-of-Thought Alignment
来源日期（仅到月份） 2026-07
来源记录日期 2026-07
雷达观测 2026-11-01 21:52
摘要
To address the increasingly severe safety risk of large language models (LLMs), reasoning-based safety alignment methods have emerged. These methods overcome the limitations of ’shallow alignment’ by exposing the model’s Chain-of-Thought (CoT), enabling auditability of safety reasoning process through both training-phase supervision and post-generation verification. However, this transparency creates a critical vulnerability, a tension we define as the Security Auditability Dilemma: while explicit reasoning is a prerequisite for safety, its textual Auditable paradoxically transforms it into an optimization target for adaptive attackers and induces the model to unintentionally copy harmful content from its own reasoning context. To address this, we propose Auditable Latent CoT Alignment (ALCA), a framework that decouples internal reasoning from external output. ALCA shifts the safety deliberation process into a continuous latent space. This allows the safety reasoning process to guide the generation of harmless outputs, while eliminates the discrete textual surface that facilitates internal copying and adaptive attack. Yet, this process is not a black box. we introduce a restricted Self-Decoding mechanism that allows the model to reconstruct its latent reasoning into human-readable text for supervision under specific guidance. Extensive experiments show that ALCA achieves robustness alignment, reducing the success rate of adaptive jailbreak attacks by over 40% compared to strong baselines, while preserving performance. Our framework presents a path toward building LLMs that are both robustly secure and auditable.
论文发现出处
这些页面提供论文线索，不代表同行评审或质量认证。
观测到的版本
平台记录号代表本站观测到的元数据变化，不是作者修订号。来源日期仅到月份或年份时，按该月或该年首日存档。
记录 #1 · 2026-07 · 最新观测版本
Resolving the Security-Auditability Dilemma with Auditable Latent Chain-of-Thought Alignment
To address the increasingly severe safety risk of large language models (LLMs), reasoning-based safety alignment methods have emerged. These methods overcome the limitations of ’shallow alignment’ by exposing the model’s Chain-of-Thought (CoT), enabling auditability of safety reasoning process through both training-phase supervision and post-generation verification. However, this transparency creates a critical vulnerability, a tension we define as the Security Auditability Dilemma: while explicit reasoning is a prerequisite for safety, its textual Auditable paradoxically transforms it into an optimization target for adaptive attackers and induces the model to unintentionally copy harmful content from its own reasoning context. To address this, we propose Auditable Latent CoT Alignment (ALCA), a framework that decouples internal reasoning from external output. ALCA shifts the safety deliberation process into a continuous latent space. This allows the safety reasoning process to guide the generation of harmless outputs, while eliminates the discrete textual surface that facilitates internal copying and adaptive attack. Yet, this process is not a black box. we introduce a restricted Self-Decoding mechanism that allows the model to reconstruct its latent reasoning into human-readable text for supervision under specific guidance. Extensive experiments show that ALCA achieves robustness alignment, reducing the success rate of adaptive jailbreak attacks by over 40% compared to strong baselines, while preserving performance. Our framework presents a path toward building LLMs that are both robustly secure and auditable.
查看此版本解读 →查看此版本原文 ↗