D squared-Monitor: Dynamic Safety Monitoring for Diffusion LLMs via Hesitation Signals
cs.AI
Submitted: 2026-05-25
Updated: 2026-09-30
Code: https://github.com/meta-llama/PurpleLlama
License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/
Sources
- GPT-4 Technical Report
- Understanding intermediate layers using linear classifier probes
- LLaDA2.0: Scaling Up Diffusion Language Models to 100B
- LLaDA2.1: Speeding Up Text Diffusion via Token Editing
- Evaluating Large Language Models Trained on Code
- The Alignment Curse: Modality Alignment Supercharges Audio Attacks via Text Transfer
- Training Verifiers to Solve Math Word Problems
- Constitutional Classifiers++: Efficient Production-Grade Defenses against Universal Jailbreaks
- The Llama 3 Herd of Models
- Safer by Diffusion, Broken by Context: Diffusion LLM's Safety Blessing and Its Failure Mode
- Llama Guard: LLM-based Input-Output Safeguard for Human-AI Conversations
- Mercury: Ultra-Fast Language Models Based on Diffusion
- Building Production-Ready Probes For Gemini
- GuardReasoner: Towards Reasoning-based LLM Safeguards
- A Holistic Approach to Undesired Content Detection in the Real World
- The Attacker Moves Second: Stronger Adaptive Attacks Bypass Defenses Against Llm Jailbreaks and Prompt Injections
- Granite Guardian
- DLM-Scope: Mechanistic Interpretability of Diffusion Language Models via Sparse Autoencoders
- Qwen3 Technical Report
- ShieldGemma 2: Robust and Tractable Image Content Moderation
Related papers
- MAVEN-T: Reinforced Heterogeneous Distillation for Real-Time Multi-Agent Trajectory Prediction
- Model Discovery Agent: LLM-assisted Bayesian experiment design for data-efficient discovery of mechanistic world models
- Agents' Last Exam
- The Clinician's Veto: Navigating Trust, Liability, and Uncertainty in Autonomous AI Prescribing
- MindHelper: Closed-Loop Embodied Mental-State Reasoning for Precision Intervention
- Incumbent Advantage: Brand Bias and Cognitive Manipulation Dynamics in LLM Recommendation Systems