Safety Reconstructed: Generative Modeling via Masked Diffusion Builds Strong Safety Guardrails

arXiv:2609.33634 · cs.CL, cs.AI · Submitted 2026-09-27 · Read on arXiv

cs.CL, cs.AI

Submitted: 2026-09-27

Updated: 2026-09-27

Code: https://github.com/meta-llama/PurpleLlama

Terminology

• Calibration — 23× in this paper · explained in 7 episodes
• Benchmark — 20× in this paper · explained in 2 episodes
• Diffusion Model — 9× in this paper · explained in 8 episodes
• Positional Bias — 7× in this paper · explained in 2 episodes
• Accuracy — 6× in this paper · explained in 1 episode
• Ablation — 5× in this paper · explained in 1 episode
• Diffusion Models — 5× in this paper · explained in 6 episodes
• Alignment — 4× in this paper · explained in 1 episode
• Fine-Tuning — 4× in this paper · explained in 5 episodes
• Failure Modes — 3× in this paper · explained in 3 episodes
• Generative Modeling — 3× in this paper · explained in 3 episodes
• Robustness — 3× in this paper · explained in 5 episodes
• Adversarial Attacks — 2× in this paper · explained in 1 episode
• Distribution Shift — 2× in this paper · explained in 6 episodes
• F1 Score — 2× in this paper · explained in 3 episodes
• Fidelity — 2× in this paper · explained in 2 episodes
• Generalization — 2× in this paper · explained in 13 episodes
• International Conference on Learning Representations — 2× in this paper
• International Conference on Machine Learning — 2× in this paper
• Adam — 1× in this paper · explained in 2 episodes
• Adaptation — 1× in this paper · explained in 1 episode
• Area Under the Precision-Recall Curve — 1× in this paper
• Core — 1× in this paper · explained in 2 episodes
• Distillation — 1× in this paper · explained in 2 episodes
• Generative AI — 1× in this paper · explained in 2 episodes
• Large Language Models (LLMs) — 1× in this paper · explained in 6 episodes
• Multi-Agent System — 1× in this paper · explained in 6 episodes
• Reinforcement Learning — 1× in this paper · explained in 4 episodes
• Sample Efficiency — 1× in this paper · explained in 4 episodes

Sources

• OR-Bench: An Over-Refusal Benchmark for Large Language Models
• Safe RLHF: Safe Reinforcement Learning from Human Feedback
• WildGuard: Open One-Stop Moderation Tools for Safety Risks, Jailbreaks, and Refusals of LLMs
• An Overview of Catastrophic AI Risks
• LoRA: Low-Rank Adaptation of Large Language Models
• Llama Guard: LLM-based Input-Output Safeguard for Human-AI Conversations
• BeaverTails: Towards Improved Safety Alignment of LLM via a Human-Preference Dataset
• CultureGuard: Towards Culturally-Aware Dataset and Guard Model for Multilingual Safety Applications
• PolyGuard: A Multilingual Safety Moderation Tool for 17 Languages
• Generative Classifiers Avoid Shortcut Solutions
• On Calibration of LLM-based Guard Models for Reliable Content Moderation
• Lost in the Middle: How Language Models Use Long Contexts
• HarmBench: A Standardized Evaluation Framework for Automated Red Teaming and Robust Refusal
• Noisy Channel Language Model Prompting for Few-Shot Text Classification
• Large Language Diffusion Models
• Granite Guardian
• Attention Sorting Combats Recency Bias In Long Context Language Models
• "Do Anything Now": Characterizing and Evaluating In-The-Wild Jailbreak Prompts on Large Language Models
• Jailbroken: How Does LLM Safety Training Fail?
• Ethical and social risks of harm from Language Models

Related papers

• Exploring Solution Divergence and Its Effect on Large Language Model Problem Solving
• Ishigaki-IDS-Bench: A Benchmark for Generating Information Delivery Specification from BIM Information Requirements
• Subliminal Steering: Stronger Encoding of Hidden Signals
• MedStruct-S: A Benchmark for Key Discovery, Key-Conditioned QA and Semi-Structured Extraction from OCR Clinical Reports
• The End of Transformers? On Challenging Attention and the Rise of Sub-Quadratic Architectures
• Untangling the Mechanisms of Misleading Context in Medical Question Answering