Hidden in Plain Sight: Advancing Harmful Content Understanding via Perception-Aware Adversarial Self-Play
cs.AI, cs.CL
Submitted: 2026-10-11
Updated: 2026-10-14
Sources
- Think you have Solved Question Answering? Try ARC, the AI2 Reasoning Challenge
- ChatGLM: A Family of Large Language Models from GLM-130B to GLM-4 All Tools
- The Llama 3 Herd of Models
- Llama Guard: LLM-based Input-Output Safeguard for Human-AI Conversations
- EvoHarmBench: Breaking Content Moderation with Iterative Human-Like Evasion
- Disentangling Intent from Role: Adversarial Self-Play for Persona-Invariant Safety Alignment
- Learning Robust Reasoning through Guided Adversarial Self-Play
- DeepSeek-V3.2: Pushing the Frontier of Open Large Language Models
- ChineseHarm-Bench: A Chinese Harmful Content Detection Benchmark
- Chasing Moving Targets with Online Self-Play Reinforcement Learning for Safer Language Models
- DeepSeekMath: Pushing the Limits of Mathematical Reasoning in Open Language Models
- GLM-4.5V and GLM-4.1V-Thinking: Towards Versatile Multimodal Reasoning with Scalable Reinforcement Learning
- MAGIC: A Co-Evolving Attacker-Defender Adversarial Game for Robust LLM Safety
- S-GRPO: Unified Post-Training for Large Vision-Language Models
- Qwen3 Technical Report
- RigorLLM: Resilient Guardrails for Large Language Models against Undesired Content
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