Unread or Unenforced? Separating Representation from Enforcement Failure in Content Guards

arXiv:2609.26178 · cs.CR · Submitted 2026-08-12 · Read on arXiv

cs.CR

Submitted: 2026-08-12

Updated: 2026-09-24

License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/

Terminology

• Decode — 70× in this paper · explained in 2 episodes
• Linear probe — 5× in this paper · explained in 2 episodes
• Failure Modes — 3× in this paper · explained in 3 episodes
• Fine-Tuning — 2× in this paper · explained in 5 episodes
• Ablation — 1× in this paper · explained in 1 episode
• Alignment — 1× in this paper · explained in 1 episode
• Benchmark — 1× in this paper · explained in 2 episodes
• Distribution Shift — 1× in this paper · explained in 5 episodes
• Generalization — 1× in this paper · explained in 13 episodes
• Robustness — 1× in this paper · explained in 5 episodes
• Trace — 1× in this paper · explained in 2 episodes
• Transformer — 1× in this paper · explained in 2 episodes

Sources

• When Refusal Looks Safe: The Refusal-Cue Shortcut in Safety Guard Models
• The Llama 3 Herd of Models
• Llama Guard: LLM-based Input-Output Safeguard for Human-AI Conversations
• LLM Safety From Within: Detecting Harmful Content with Internal Representations
• Constitutional Classifiers: Defending against Universal Jailbreaks across Thousands of Hours of Red Teaming
• DecipherGuard: Understanding and Deciphering Jailbreak Prompts for a Safer Deployment of Intelligent Software Systems
• Towards Safeguarding LLM Fine-tuning APIs against Cipher Attacks

Related papers

• SoK: AI-Augmented Binary Reversing
• Relaxed Sender Anonymity for CBDC Interbank Settlement: A Zero-Knowledge Approach on Permissioned EVM
• Calibration-Family Overfit: Why Trusted Sabotage Monitors Don't Transfer Across Lineages
• Efficient Fuzzy PSI under One-Sided Assumptions
• Sealing the Audit-Runtime Gap for LLM Skills
• Token Composition: A Graph Based on EVM Logs